Conference heron::euro_swas_ai

    Joe,
    
    I understand that in these difficult times, Geneva people try to 
    track every unuseful/unprofitable activity. Is AI unprofitable? Is 
    AI CC's activity unuseful?
    
    I hope the following updated report I submitted to Ed Orciuch last 
    april on French AI activity will bring some of the pieces of 
    evidence you're looking for.
    
    What it shows is :
    1)	AI Business exists; if direct Digital project revenue is still 
        too low, partners generated business is significant; projects 
        where Digital is directly implied is about to generate 
        important leverage business (account penetration), 
        unfortunately difficult to measure ;
    
    2)	Many presales contacts underway. Very interesting leads have 
        been collected from Avignon 91 (Shell, Kodak, UAP, Alcatel, 
        Minist�re des Transports...
    
    3) 	BUT... there are too few specialists to manage both ongoing 
        projects and long presales activity : much of the AI business 
        goes to partners or to competion (SOLLAC's SACHEM 80 MFF black 
        furnace project won by Bull/CAP Sesa)
    
    4) 	Training issues : Projects often include demand for specific 
        training. We've had a problem to have those needs satisfied by 
        EDU services who want to stick to catalogue training offers 
        quite inadequate. As a consequence we had to give courses by 
        ourselves. I guess that other countries might have experienced 
        the same kind of situation.
    
    Conclusions
    
    We must have a way to maintain (dicreasing) AI competency, 
    stabilize it, and why not increasing it : a market is waiting for 
    that. If we don't go into AI project business, the competition 
    will grab it (software houses, BULL, Siemens...).
    
    I think the directions drawn during last october CC meetings are 
    the good ones (ie. AI awareness on the field, AI and project mgt, 
    extend capabilities from E.S. to the so-called "emerging" 
    technologies, organize customer seminars).
    
    This is critical to develop more country AI skills. And we need to 
    have more in order not to be overwhelmed by both presales and 
    project. But perhaps more s/w engineers with AI culture than "pure 
    AI" specialists.
    
    Last but not least, YES we need AI services marketing support (up 
    to now we've been our own...) in order to get a structured 
    services portfolio, promote it internally/externally and sell it.  
    
    These I thinks are a few good reasons (there might be many more) 
    for undoing AI-CC delete.
    
    ==================================================================
    		
    		Some infos on French AI activity FY91
    
       
    AI projects :
    
      -	CIT-ALCATEL AI Fellowship (expert telephone network 
        monitoring) ; 
      -	COM'AZUR expert server in operation since last october.
      -	ROQUETTE (Planning system for food industry) : small 
        maintainance to be done this year (200-300 KFF) ;
      -	VALLOUREC expert BOM generation : prototype presented in may ;
      -	SOCS project : conversion of banking application from 
        VM/PROLOG to PROLOG.
    
      
    
    Pre-sales activity :
      -	COGEMA : screening of computer user calls using expert systems 
        and voice interface ; process diagnosis and monitoring to be 
        discussed ;
      - ISL : AI Fellowship to start by the end of calendar year ;
      -	Lille Hospital : toxicologic problems monitoring system ; 
        proposal for a feasability study ;
      -	DCN : expert systems feasability proposal for managing 
        Technical Instructions ;
      -	PROST : interest for AI and transport ;
      -	Peugeot : meeting planned with Technology Mgmt;
      -	ASF project managed by Valbonne ;
      -	3rd parties selling projects for Aeroports de Paris and 
        Martell, Cr�dit Agricole, CCF, Cr�dit Lyonnais...) ;
    
    Business status  for projects for which we have a direct 
        visibility :
      - direct Digital projects : 1.5 MFF ;
      -	close 3rd parties : more than 2.0 MFF s/w 
      -	many projects can't be done/are lost due to late implication 
        in the selling cycle (e.g. Schlumberger, Soretex, Airbus 
        Consortium Order Mgt system, ...) ;
      -	projects made by Digital France often subcontracted ;
      -	many projects "left" to 3rd parties without control, for 
        lack-of-resources reasons.
    
    French Competency Circle Activity :
      
      -	seminar Digital/AI corp/ILOG in january (25 people);
      -	initiated press communication & writing of AI selling guide 
        for sales & EIS people ;
      -	Competency Tranfer Programs ;
      -	Avignon 91 participation.
    
    Problems & Support required :
    
      -	need more follow-on of European AI Competency Circle meetings 
        (cf last october actions on DPM tuning for ES projects, 
        country teams, information of account managers) ;
      
      -	participation in Circle meetings is low ; 2 leading 
        individualities lost (1 has left, 1 has got a different job) ;
      
      -	need more coordination on presales activity in France directly 
        managed by EIC/AI Valbonne, especially in the following 
        regions : South, Paris Finance (Puteaux), Paris Industrie 
        (Antony) ;
      
      -	although France appreciates very much cooperation with 
        Valbonne teams, the effect of scattered action is perverse on 
        our goal to "bring AI to the field" and yields just the 
        opposite of what expected last october : 
      
      	. getting used to Valbonne's action French EIS management of 
          at least 4 Regions of 5 is no longer motivated to invest in 
          AI so that still alive AI specialists are "cannibalized" by 
          short-term business activity;
      
      	. EIC/AI comes to projects not correctly qualified by the 
          field on business aspects, with induced risk to invest time 
          and sweat inappropriately.
      
      -	account teams sometimes directly involve Valbonne or US teams 
        on actions without local visibility.
    
      -	what we suggest /need :
      	. coordination with Valbonne teams for project on a regular 
          basis (French AICC meetings for instance), to define & 
          dispatch responsabilities (presales, project) ;
      	. systematic mutual information on French projects ;
      	. support to promote AI internally and outside ; participation 
          of Valbonne to our "AI selling guide" would be appreciated ;
      	. support to bring AI culture to EIS specialists & project 
          managers.
      
      =================== End of report ==========================

Joe, thanks for taking the issue! It's good to know there is still White Knight
who's fighting the Evil!

I fully agree with Stephane in .1
Here are my $.02 Joe. 

Luc

-------------------------------------------------------------------------------
>>>>>>>	- AI related project activity in European countries and revenues gained

There is no local AI-cc in Belgium. The technology is covered in the CASE group,
merely because the AI-people are part of the CASE Entrepeneur Support Unit.
This works fine in Belgium, since we position AI (or ES) as Software Engineering
using other than conventional techniques.

For several occasions this approach, and the ability to talk about that 
technology, opened (or helped opening) door.(Thanks to the Valbonne-AI group for
you active support!)
We could show that we could handle or understand the customer's problem, 
and are able to solve it in several ways. 

We (digital Belgium) have not realised
any customer project with AI-contents yet. The reasons for this :
	- limited local support (1.5 people)
	- distance to Valbonne
	- AI techniques helped to better understand problem,
          which eventually could be solved the traditional way.

But we do have customers which are active or considering using the technology.
Some names? :
	- SIDMAR (steel factory)
		- continuous research for better production methods

	- COCKERILL (steel factory)
		- process monitoring

	- Forges de Glabecq (steel factory)
		- production planning

	- Picanol (machine builders)
		- fault diagnosis

	- Phillips 
		- planning

	- BIKIT (University research)
		- several customer projects

	- SWIFT (banking network)
		- continous research

	- ...

Besides the customers, some Solution Providers are very active in the Belgian
market :
	- AICORP (KBMS)
	- GENSYM (G2)
	- AI-SYSTEMS (G2, KBMS, Nexpert, projects)
	- Software Generation (ADS)
	- E2S (projects)

Why not Digital? See Marketing support


>>>>>>>	- AI related training..........

Chicken and Egg problem :
	as Stephane already mentioned : in order to be able to do some projects
	one needs special knowledge, for which no training is offered.
How to overcome?
	Maybe the experienced people (non-edu) can train others on when needed
 	basis? or assist in project start and thus deliver on-the-job-training.
	Promotion of internal AI-projects.


>>>>>>>	- AI-CC activity over last 18-24 months, your perceived benefits.
	benefits for my daily job :
		-technology update (is this training?)
		-experiences exchange, technical and commercial (is this training?)
		-expanding personal network of internal 'solution providers'
		- ...

>>>>>>>	- general direction which the AI-CC should take, expected benefits, 
>>>>>>>	  your perceived impact of no AI-CC ("Delete" function applied!)

the AI-CC was going in the right direction !
	
	1) awareness.
	During the last meeting the AI-awareness problem was discussed and
	possible solutions were raised.
	BUT WERE ARE THE ACTIONS??? (see marketing issue to)
	The fact that WE, the field have to come up now to convince GEO,
	proves that there were no convincing actions undertaken since last
	meeting.

	2) contents.
	imho I think the AI-cc should widen the scope a little bit.
	(as it tended to do during the last few meetings)
	We should cover some emmerging technologies like OO, MM.
	I know this is a large domain, but to reduce it, I propose to
	cover those topics/technologies/products/... which are usable,
	sellable, projectable, ... today or in the very near future.

>>>>>>>	- Is there a need to develop more, any?!, AI skills in European 
>>>>>>>	  countries, where, why would you expect benefits.

	YES, YES, YES, ....
	BUT! not as a standalone action. The purpose is not creating highly
	trained specialist to WAIT for customer questions.
	Training should be accompanied by marketing actions.

	Benefits? : pff, only sales.

>>>>>>>	- Do you feel a need for a AI services Marketing support?

Good question. It implies there is/was no Marketing support.
And that's exactly what was missing. We all raised some good ideas,
we even had a marketeer at one time, some people went to see the countries, 
everybody agrees on actionplans, and then we all sit back and WAIT.
Sorry for the flame, and I'm not aiming at somebody in particular, because 
I know everybody is trying hard to do a good job.

We (dec) are not seen by the market as an AI-able company. Everybody knows XCON,
(who did that marketing??), but no-one considers DEC as AI-partner.
Neither as OO-partner or MM-partner. WHY? Marketing!

Today we're up against things like the OO knowledge HP agressively demonstrates 
or the AI-ability IBM, BULL and others claims.

In the Emmerging Technology area we need Internal AND External marketing SOON,
or we'll loose (are loosing) market share.

So, closing the AI-cc and related function will NOT save any money, it'll COST

$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$

 

    Joe,
    
    I think it's really unuseful and dangerous to continue to speak only
    about AI. AI is only an important part of a set of technologies who
    REALLY fit the customer intereset. But.. alone, it doesn't absolutely
    satisfy both customers and Digital structures.
    
    It's perfectly stated from a long time that will be the SW the next
    hero of the Computer Science market. That's because after the dramatic
    evolution of the HW all do expect the same for the SW.
    
    For this reason ALL and GLOBALLY the Emerging technologies like OO
    Technology, Neural Nets, Hypermedia, and .... of course, Artificial
    Intelligence, are considered the boosters capble to give the right push
    to the new SW generation.
    
    I can understand Geneva. What's the business realted to AI? Really
    small! It doesn't absolutely justify the  investments done in this
    area! The Italian Association on AI is loosing more and more
    subscribers... while the New association I've set up in Digital (is a
    SIG DECUS about all the new SW technologies) actually has more than 300 
    subscribers (in other words more than the mentioned Italian Association). 
    Why this? because the customer has enough to listen only AI AI AI. 
    AI has been too much over-evaluated from Universities ande described 
    like the solution for all the problems. Despite there've been success 
    stories about AI, there've been also many failures caused by  the fact 
    that AI has been used to solve problems that wasn't able to perform. 
    That's the reason today the customer interest is focused on those
    technologies who could complement AI in solving those problems that the
    new HW performances allow.
    
    So.... let's stop with AI CC and let's start with EAST CC (Emerging
    Advanced Software Technologies). Future European Business is in EAST
    (says also Gorbachew  ;-)  )
    
    About Italian market I can tell you many success stories but they don't
    concern only AI.
    
    Here some of them:
    
    ENEA
    	IDEA Project: Use of Neural Nets, AI, OO Technology for three
    	important research streams: 
    
    	1)	Robotics and Vision
    	2)	Advanced information management and Office Automation
    	3)	Environment management.
    
    	LEVERAGED HW: VAX 9440 (largest ever sold by the Corporate) that
    			vill be used for the development of the projects
    
    	LEVERAGED CONSULTANCY: 36 MAN/YEARS for the developmentof these
    			projects.
    
    BONIFICA:	
    	RASVECTOR Project: Using Neural Nets and Object Oriented database
    	to exctract manage and save information from satellite photos
    	for mapping.
    
    	LEVERAGED HW: 2 * VAX9410 (One devoted to the project development)
    	
    DEFENSE MINISTRY: Top Secret Project using Neural Nets. HIGHLY
    	STRATEGIC for Digital Italy. Sorry! I must not provide information
    	about this project.
    
    ITALCABLE: Two Project concerning thew use of Nueral nets to develop an
    	information management system for their manual (consultancy amount
    	about 200K$)
    
    PIRELLI - WHIRLPOOL: Expert system routinely used by Whirlpool for the
    production scheduling of European frigiders.
     
    

    Going on with the project:
    
    GD: 
    	DAYSY Project: AI has helped us in cracking in an IBM shop.
    Actuallly, after havig sold 20 workstations, some PDP and servers,
    Digital, showing to be the most advanced technology partener using AI,
    is going to sell a 9000.
    
    POLICE: Two starting project:
    
    	The first is TOP SECRET and cannot be described (will use Neural
    Nets) but is considered highly strategic. The second is a training system 
    for the Police School (probably based  on Merlin)
    
    FINANCE POLICE:
    
    	Starting Project to develop an  Expert System to help policemen in
    controlly if the Companies avoid to pay the taxes.
    
    I would to stress that all the projects involving Defense Ministry and
    Police are for us very important because are a powerful tool to go into
    the power centers of the Government.
    
    ITALIAN BANK (CRT Turin): Two projects using Expert systems (one ended
    and routinely used) for the Credit analysis.
    
    ALFA AVIO: Startig project to develop an Expert System for failure
    repairing of Elicopters engines. This poroject for us is very strategic
    to leverage the business concerning the realization of the second
    Factory of Alfa Romeo Avionics (business leveraged about 4 Million $).
    
    These are only the most ipmortant I've now in mind. If you want more
    detailed information, in particular about top secret projects, please
    contact me directly by phone (793-8490)
    
    Bye
    
    Ricccardo
    

Dear all,

        Here's my tuppence too...
        I fear I've really flamed-on a bit below, but I'm d***d if I've done
        all this typing for nothing.  As the great Dylan once said

                "I've suffered for my music....

                                ...now it's YOUR turn."


        Joe's original points are marked with ** (The order is changed)

        Also, please forgive the Gulf-War rhetoric, I got carried away...


** General direction which the AI-CC should take, expected benefits, 

        I think the CC could be a Pan-European "Rapid Reaction Force" of
        say 4-5 in large countries, and 1-2 in small countries.  These
        people should be good SW engineers generally, have a solid
        grounding in AI and also PRESENTATION/EDUCATION SKILLS to pass on
        their experience to local project teams.

        These numbers should be maintained at all times, that is
        *replenished* if necessary (Use the VBO AI trg string).  Thus we
        can pick up on business (that DOES exist and is going to our
        competitors) and draw in reinforcements from the rest of the CC if
        necessary to ai-train (transfer technology to) local personnel on a
        project-specific basis.

        The CCMems should know each other (personally) and know of
        each-other's work.  Thus we need the TEC-Fora/CC meetings.  We also
        need to use and feed the AI-projects database (where is it now ?)

        There should be liaison with similar groups in USA and GIA.


        In anticipation of the suggestion that the activities described
        here are precisely what the already-established Valbonne groups are
        doing, I'd like to point out a couple of differences:

                -the expertise is perhaps not so high (for CC members), but
                 more distributed and accessible to "grass-roots" 
                 account-teams etc.  They should be less expensive too :-) ?

                -CC members bringing in more experience of working in
                 customer-projects for extended periods

                - Not only are they (the CCMems) more available to the
                 "grass-roots" for output, they are closer for *input*. 
                 That is, they know, and they report regularly (at
                 TEC-Fora!),  on trends and developments in their own
                 countries/ fields of  particular expertise.

                - The emphasis on short, precision-training based on personal
                  experiences in the field (see below).



** AI related training

        In my (all too personal) experience, AI is too low-demand for EDU. 
        No provincial trg centre (eg Munich) can afford to keep a half-time
        AI trainer (let alone *full* time).  Having said that, trainers in
        this instance are probably *not* what's required...

        ...AI trg should be given if possible by CC members with relevant
        *experience* to the (potential) project at hand.  I suggest that CC
        SW-engineers who have used AI in a project be given AI-CC funded
        time to put together a 1-2 day mini-course (say 20 slides) on
        lessons learned/tools used in that project, and that they be
        available on request to give that presentation (+/- a day or two
        consulting) anywhere in Europe.  This activity to be tracked and
        coordinated by the CC.

        Should maybe concentrate on the *added value* AI techniques can
        give to SW.  AI trg should be given to already good SW engineers.
        
        Tool trg can be given late and should be tailored to the project
        domain (by examples) if possible.

        
** Your perceived impact of no AI-CC ("Delete" function applied!)

        Drifting apart of, and lack of coordination between AI-capable 
        SW-engineers in Europe (that magnificent and rare breed!)
        
        hence...
        
        Re-(re-)invention of modules and demos which might swing a project
        DEC's way (NOTHING sells a project better than the
        line "Oh, we've done that before in a project in ___ !" )

        hence... lost time

        hence... lost business

	hence... lost $$$  
    
    
** Is there a need to develop more, any?!, AI skills in European 
          countries, where, why would you expect benefits.

        (see above) Decide upon and KEEP UP the minimum numbers for the 
        AI-CC "Rapid Reaction Force".  These people should then do
        pin-point, project-point "precision bombing" (I'm really getting
        carried away here!)  - that is technology-transfer to bring local
        proj-teams up to speed in AI techniques as fast as possible, as and
        when the business arises (so they don't have time to forget what
        they learn before they need it!)


** Do you feel a need for a AI services Marketing support?

        As Luc points out , we are not perceived by customers as a seriously
        AI-capable company.  Why not ?

        It is also important that we, the field, perceive *ourselves* to be
        an AI-capable company, that the field knows that the AI-CC exists
        and is able to deploy (here I go again) people to help them.


** AI related project activity in European countries and revenues gained

        For the past year+� I've been directly involved in an AI project at
        Sandoz in Basel, which was acquired by Caroline Stutz (formerly
        ACT-AI Zuerich(see NOTE below)) on a *purely AI* basis.

        Sold to date:
        1 Analysis and Concept level Design
        1 VS 3100
        1.5 MYears paid consultancy

        Other Benefits:
        Visibility in a strongly IBM-minded company

        NOTE:  1-jul-91 DEC-Switz. disbanded the ACT and all AI activity,
        with exception of the one guy still working at Sandoz, has stopped.
        DEC no longer considers itself "ready for AI" in CH!

        AI in Germany has been concentrating on "niche products" with AI
        content (eg ATEM and UFIS) for which there is a HIGH demand in
        terms of consultancy.  We will lose the projects if there
        is always a little SW-house down the road that can do the job with
        half our overheads (usually after we've invested big $$$ on
        presales efforts!) UNLESS we can prove that "We've done it before".

        

The preceding entry (-.1) is - in it's final form - intended for the
 European CC Co-ordinator, Subir Choudhury, and his boss Lennart Hjord who 
is the European Sales Support and Technology Manager. I will correlate all 
your inputs on my return from Vacances (20th Aug.) Until then please read 
and feedback your comments/inputs.

Thanks,
Joe.

    
    Joe,
    
    Looks Ok.  Could be a bit fleshier in the areas of Mktg and Trg needs.
    
    
    re.p 13:  
     "Currently the largest European AI projects with exclusive Digital
    staff are run in Switzerland"
    
    As far as I know, Sandoz is the only AI proj in CH, and that is under
    Sandoz control (As of last week only 1 DEC SW engineer participating,
    in a team of 2-3)
    
    Sorry, but I think Geneva would have noticed!
    
    -pete

    Joe will be taking these inputs/our case to the relevant people
    in late August. Bearing this in mind, I suggest we anticipate
    postponement at least of the planned September date; it doesn't
    give us enough time bewtween (favourable) decision and date
    to organise properly. So, hopefully it will still be on, but at
    a later date.
    
    Steve

The following was compiled at the request of NAS marketing to support their 
advertising campaigns...

Pat

p.s. I would be very careful about throwing away the AI title. We are on the
verge of introducing to industry tools like IMKA, KARMA, Symbolic Modeling, 
Market Data Navigator, Fianacial Navigator, ECCAP and other tools and 
applications that are starting to deliver on the inflated promises that people
have tied to AI over the years. Very senior Digital management in the US is 
starting to catch the fevor as they are starting to see the impact on our own
business and the implication that taking it to our customers has. It is also
being seen by their European counterparts. Now that we are positioned to 
deliver, I would hate to see us throw confusion in the game by equating AI only
to expert systems and trying to bury it as a subset of emerging technologies.

-

A couple of other sources of support for continued existence of the AI CC are 
two reprints of magazine articles that can be requested from 
Adam (SELECT::) Couture...

Cover article from the June 3, 1991 issue of Fortune Magazine titled "BRAIN
POWER - How Intellectual Capital Is Becoming America's Most Valuable Asset"

From the State of the Art section of Byte Magazine, January 1991 titled 
"AI IN PRACTICE - Once it works, does it still qualify as AI?" - mostly about 
DEC and basically answers the question by saying, "Who cares, as long as it 
works and solves problems that could not have been solved before.

If you need a copy REAL quick, send me a note. I have about 6 copies of each. Do
it quick though. If you haven't yet heard (I just recently found out myself), my
family and I are relocating back to the US at the end of August. I will be going
back to work in the AI Technology Center. I'll let you know what I'll be doing 
there once we finalize exactly what it is! 

Please know that I truely value the relationships that I have built over the 
years with all of you and will miss working with you directly. I certainly plan
to try and find every opportunity to continue a personal and professional 
relationship with the European AI community. Please feel free to consider me as
a "US" resource.

Thanks again for your freindships and all you have taught me over the years.

Pat

================================================================================



Mike,

Here is far mor information on AI than you ever wanted to see.

For software developers the important points to emphasize are:

1.  Digital's recognized leadership in the field of Artificial Intelligence.

2.  All of our solutions, tools, training and consulting are consistent

with Digital's commitment to the NAS architecture.

3.  We can provide tools, training and solutions to third parties to develop

their solutions.  Most notably, training in "Knowledge Based Methodologies" and 

the "Fastrak" program would be applicable to third party solution developers.

4.  The areas of application concentration for Digital's Artificial
Intelligence Technology Center are:

	A.  Concurrent Engineering
	B.  Knowledge-Based Information Retrieval
	C.  Intelligent Computer Systems Managment
	D.  Large Scale Configuration
	E.  Business Planning and Modelling

(Ed/Pat more detail on these from business unit plan or AAAI materials)

Following is verbeage which should help considerably.  At the end I've also
listed some of the DEC and third party products and the Digital internal
contacts for them.

Give me a call at DTN 296-5166 if you have any questions.

Thanks,

Adam

================================================================================

Digital and Artificial Intelligence

Digital is the recognized leader in the successful implementation of AI
applications throughout the company.  Today, we have over 70 AI applications
in use worldwide, used by over 10,000 people and saving the company $200 
million dollars.  Because of this experience, Digital, more 
than any other company, is in a position to share that wealth of knowledge and 
experience.  In fact, we have established Artificial  Intelligence Technology 
Centers around the world to assist customers and third parties in employing 
Artificial Intelligence technologies as part of their overall solution.  The 
largest is in Marlboro, Massachusetts, the second biggest in
Tokyo, Japan and we also have centers in Europe and Australia.

Digital established the what we call "the knowledge network architecture"
in the early 80's to deal with complex enterprise-wide problems.  The 
knowledge network serves as a blueprint for development of systems and as a 
statement that knowledge needs to transcend functional organizational barriers
to be effective.  The underlying strengths of Digital's open computing 
architecture made the implementation of the knowledge network possible.

This isn't a new concept at Digital.  It's a logical extention of our Network 
Application Support (NAS) environment which incorporates a wide variety of 
hardware and software technologies, integrating a multivendor environment in a
transparent and productive manner.  The same way that NAS integrates the 
systems of an enterprise, DECwindows integrates the user interface and access 
capabilities of end users.  As a result, the learning curve for using new 
tools and applications is minimized and cost savings and effectiveness 
through knowledge based solutions is maximized.

This open advantage strategy has is driving us to a very unique and focused 
AI product strategy . . . one that embeds the technology in our products as 
well as our internal processes.  For our customers, it means that we must 
also provide leading AI products as well as complete training and consulting 
services.

We're implementing this strategy by assuring that AI technology is incorporated
in the mainstream of solutions for our customers, our third parties and, of 
course, our own internal processes.  Of course, this means supporting standards 
like NAS, OSF, VMS and Unix.

To help our customers and third parties implement knowledge-based technology
into their solutions, we are constantly updating our wide variety of products
and services including consulting services, training, and our industry-leading
fellowship program.

FELLOWSHIP PROGRAM

THE DIGITAL Business Fellowship in Artificial Intelligence is an intensive

six-months residency training program offered by Digital to share our knowledge

and experience with knowledge based technology.  The program includes a

two-and-a-half month core curriculum of structured classroom and hands-on

laboratory training, followed by a four-month apprenticeship supervised

directly by senior Digital knowleldge engineers.  The result is a competently

trained knowledge engineer who can then apply knowledge based systems to meet

specific business needs within his or her own organization.


SELECT CONSULTING

SELECT Consulting transfers proven technology to other enterprises via 

management and technical training programs and consulting services that 

utilize the same software tools we use internally to build our own knowledge 

based solutions.  Utilizing standard Digital products, third-party products, 

or knowledge based systems developed for internal use, AI SELECT adapts 

software tools specifically for each customer's profile and problems.


TRAINING


Digital offers both curriculum-based training and point courses in the

knowledge based systems arena.  Many of the software products offered by both

ourselves and our third parties are supported with Digital training.  The

curriculum based training is focused at both the management and knowledge

engineer level.  Our knowledge engineering curriculum is the same one used

internally at Digital to train our own knowledge engineers.  The management

curriculum consists of two four-day modules that address the broadest

implications of knowledge based systems in an organization.  Most recently,

Digital introduced an intensive 10-day program in the essential concepts and

methods of knowledge engineering including specific methodologies necessary to

develop and implement knowledge-based system solutions.  A new one-month

curriculum providing an overview of knowledge-based sysgems technology along

with specific skills and hands-on application training necessary to apply AI

technology was also recently announced.  


AI PRODUCTS ON DIGITAL PLATFORMS

_DECscheduler_, V1.1, an application for distributed systems management.  It

is a scheudler of vax jobs designed primarily for data centers.  DECscheduler

was developed by Digital for internal use and is running at more than 900

locations and at 200 Digital customer sites.  It runs on VMS and was announced

as a Digital product in February of 1990 - DEC contact, Bonnie Morrissey

DTN 291-9443.


_KBMS_ was developed by AICorp, one of the oldest AI companies, with which

Digital has a joint marketing agreement.  KBMS is a srong MIS-oriented

knowledge based application development package with exceptional database

integration capabilities.  AICorp is a Digital Complementary Solution

House(CSH).  DEC Contact Linda Reding DTN 296-5183.

(Ed/Pat: AICorp is scheduled for CMP review 10 September 1991)


_ADS_ was developed by AION Corporation. ADS is a srong DP-oriented

knowledge based application development package with exceptional transasction 

processing capabilities.  AICorp is a Digital Complementary Solution

House(CSH).  DEC Contact Linda Reding DTN 296-5183


_LASER_ is a robust, inexpensive object-oriented rule-based tool written

in the C language by its developer, Bell Atlantic Software Systems.  LASER

runs on VAX/VMS, VAX/ULTRIX and RISC/ULTRIX.  Bell Atlantic is a Digital

CSH.  DEC Contact Linda Reding DTN 296-5183.



_Lucid LISP_ is perhaps the most popular implementation of Common LISP

available on the market today.  Lucid LISP runs on the RISC/ULTRIX platform

and is a Digital Distributed Software(DDS) product.

DEC contact  Andy Ostrom DTN 244-7821. 

_VAX LISP_ is Digital's implementation of the Common LISP standard, 

developed to run on the VAX/VMS platforms.  DEC contact  Andy Ostrom DTN 

244-7821. 


_VAX Decision Expert_, a diagnostic productivity designed for easy use

by the novice AI programmer, was developed in the C programming language

and runs on Digital's VAX/VMS platforms.  DEC contact  Andy Ostrom DTN 244-7821. 


_Nexpert OBJECT_, developed by Neuron Data, is another DDS product sold and

supported by Digital.  It is a rule-based tool with object oriented

capabilities.  Nexpert OBJECT features include an agenda monitor and multi-

valued slots.  This product runs on VAX/VMS and VAX/ULTRIX.

DEC contact  Andy Ostrom DTN 244-7821. 


_VAX OPS5_, developed by Digital in BLISS, is a forward-chaining, rule-based

implementation of OPS optimized for performance.  Most of Digital's

critical knowledge-based applications have been developed on OPS5.  It runs

on VAX/VMS platforms.  DEC contact  Andy Ostrom DTN 244-7821. 



_Knowledge Craft_  A knowledge representation and problem solving tool,

Knowledge Craft was developed by Carnegie Group, which has a CMP relationship

with Digital.  Available since 1985, Knowledge Craft is written in LISP and

well-suited to large knowledge based systems applications.  It runs on the

VAX/VMS platforms.  DEC Contact Linda Reding DTN 296-5183.



_Quintus Prolog_, also a DDS product, is Quintus' implementation of the Prolog

language with an interactive development environment.  Available since

1985, Quintus Prolog runs on VAX/VMS and VAX/ULTRIX.  DEC Contact Linda Reding 

DTN 296-5183.
 


_Rdb Expert_, the result of a five-year development program at Digital, is an

expert system that generates optimal storage design for VAX Rdb/VMS and 

VAXspace/DBMS.  It helps gather data needed for analysis, generates

executable SQL code to create the database and reports.  DEC contact Wendy

Herman DTN 381-2494.


_VAX Performance Advisor(VPA)_ is a knowledge-based interactive tool for

analyzing VAX System Performance Management(SPM) log file data.  VPA

can act as a performance consultant, providing immediate analysis of VAX

SPM performance data or as a tutor, providing the user the ability to 

interact with VPA and learn how performance of VAX/VMS systems is

achieved using a systematic performance methodology.  DEC contact Liz

Sooho  DTN 381-0908


_Level 5_ has been available since 1984 and was developed in C by

Information Builders, a Digital CMP.  A rule-based tool with excellent

FOCUS integration.  FOCUS is a database tool also from Information 

Builders.  Both run on VAX/VMS platforms.  DEC contact Joyce Traister

DTN 291-0420.


_Mercury KBE_ was developed by Artificial Intelligence Technologies in

LISP.  It is a knowledge-based system that stores information in relational

databases and can deal with large numbers of rules and objects.  Mercury

KBE adheres to CLOS standards and only runs on the VAX/VMS platform.

DEC Contact Linda Reding DTN 296-5183.


_TCS_  Developed by Carnegie Group, Text Categorization Shell is a

productivity shell that does automatic content-based categorization of

on-line text.  DEC Contact Linda Reding DTN 296-5183.


_Enterprise Expert_  Sold by Computer Associates, a CMP, this product is

a rule-based application development tool and has been available since

1984.  Written in COBOL, it is the tool for developing applications in

DECvoice.  DEC Contact Linda Reding DTN 296-5183.


_G2_  Developed by Gensysm, G2 is a real-time KBS which is well-suited to

process control applications that require managing and/or monitoring

information on a real-time basis.  Written in LISP, G2 has been available

since 1987 and runs on VAX/VMS and RISC/ULTRIX platforms.  Gensym is a

Digital CSH.  DEC Contact John Pincomb  DTN 291-0420.

    Great, lots of useful messages so far: lists of capabilities and of
    products (either packaged or via consultancy). These are good for
    summarising our capabilities.

    But, customers are beginning to ask "what is the Digital difference?" 
    (maybe it's our own fault: we introduced the term). 

    They know 
    	
     	- they can buy AI/ES etc consultancy from systems houses that 
          claim at least our experience in system-building and in many
          cases a great deal more

    	- that they can buy wide variety of tools that run on a wide variety
          of platforms, some of which are source-code compatible across those
          platforms

    	- they can get studies, training, seminars and tool-specific
    	  consultancy from independent consultants
                                 
    So... why buy any of these from us?  

    We've been trying to measure the demand for our services in the UK
    sales accounts (I've mailed every Entrepreneur and Service Centre
    manager as well as all the account managers we've worked for in the
    last two years) and many have asked "what is the differentiator that
    makes the customer prefer us". 

    All our internal customers now or will have budgets and want to buy
    (from Service Centres) pre-sales or consultancy activities which will
    have a good chance of measurable success. They are asking: why should
    I spend my money buying account support from you rather than from, for
    example, groups touting packaged VMS services.

    	Bryan

Bryan,

You make some good points. I don't have time right now to address all of the 
points, but let me give you the quick answers.

1) one of differentiations is a proven ($200 M annual ROI/ 70 working systems)
   track record

2) recognized leadership in applying technology ( AAAI'91 press / Dennis 
   O'Connor award / Byte and Fortune articles)

3) IMKA 

4) next generation IMKA tools (KARMA, Symbolic Modeling, etc.)

5) one stop shopping; i.e., you mentioned may people delivering consulting, or
   training, or tools, or apprenticeships. Some of the focused, point offerings
   are better that our offerings in that particular area (todays tools for 
   example). However, I know of no one that offers the comprehensive "one stop"
   range that we do - and it is very high quality when looked at in its 
   entirerty.

6) all of the above fits withing the overall Digital picture

I hope that these are the kinds of arguments that an account manager can relate
to.

Cheers,
Pat

%!PS-Adobe-2.1
%%Creator: DECwrite V1.1
%%+Copyright (c) 1990 DIGITAL EQUIPMENT CORPORATION.  
%%+All Rights Reserved.
%%DocumentFonts: (atend)
%%EndComments
%%BeginProcSet DEC_WRITE 1.06
/DEC_WRITE_dict 150 dict def DEC_WRITE_dict begin/$D save def/$I 0 def/$S 0
def/$C matrix def/$R matrix def/$L matrix def/$E matrix def/pat1{/px exch
def/pa 8 array def 0 1 7{/py exch def/pw 4 string def 0 1 3{pw exch px py 1
getinterval putinterval}for pa py pw put}for}def/pat2{/pi exch def/cflag
exch def save cflag 1 eq{eoclip}{clip}ifelse newpath{clippath
pathbbox}stopped not{/ph exch def/pw exch def/py exch def/px exch def/px px
3072 div floor 3072 mul def/py py 3072 div floor 3072 mul def px py
translate/pw pw px sub 3072 div floor 1 add cvi def/ph ph py sub 3072 div
floor 1 add cvi def pw 3072 mul ph 3072 mul scale/pw pw 32 mul def/ph ph 32
mul def/px 0 def/py 0 def pw ph pi[pw 0 0 ph 0 0]{pa py get/px px 32 add
def px pw ge{/px 0 def/py py 1 add 8 mod def}if}pi type/booleantype
eq{imagemask}{image}ifelse}if restore}def/PS{/_op exch def/_np 8 string def
0 1 7{/_ii exch def/num _op _ii get def _np 7 _ii sub num -4 bitshift PX
num 15 and 4 bitshift -4 bitshift PX 4 bitshift or put}for _np}def/PX{[15 7
11 3 13 5 9 1 14 6 10 2 12 4 8 0]exch get}def/FR{0.7200 0 $E defaultmatrix
dtransform/yres exch def/xres exch def xres dup mul yres dup mul add
sqrt}def/SU{/_sf exch def/_sa exch def/_cs exch def/_mm $C currentmatrix
def/rm _sa $R rotate def/sm _cs dup $L scale def sm rm _mm _mm concatmatrix
_mm concatmatrix pop 1 0 _mm dtransform/y1 exch def/x1 exch def/_vl x1 dup
mul y1 dup mul add sqrt def/_fq FR _vl div def/_na y1 x1 atan def _mm 2 get
_mm 1 get mul _mm 0 get _mm 3 get mul sub 0 gt{{neg}/_sf load
concatprocs/_sf exch def}if _fq _na/_sf load setscreen}def/BO{/_yb exch
def/_xb exch def/_bv _bs _yb _bw mul _xb 8 idiv add get def/_mk 1 7 _xb 8
mod sub bitshift def _bv _mk and 0 ne $I 1 eq xor}def/BF{DEC_WRITE_dict
begin/_yy exch def/_xx exch def/_xi _xx 1 add 2 div _bp mul cvi def/_yi _yy
1 add 2 div _bp mul cvi def _xi _yi BO{/_nb _nb 1 add def 1}{/_fb _fb 1 add
def 0}ifelse end}def/setpattern{/_cz exch def/_bw exch def/_bp exch def/_bs
exch PS def/_nb 0 def/_fb 0 def _cz 0/BF load SU{}settransfer _fb _fb _nb
add div setgray/$S 1 def}def/invertpattern{$S 0 eq{{1 exch
sub}currenttransfer concatprocs settransfer}if}def/invertscreen{/$I 1
def/$S 0 def}def/revertscreen{/$I 0 def}def/setrect{/$h exch def/$w exch
def/$y exch def/$x exch def newpath $x $y moveto $w $x add $y lineto $w $x
add $h $y add lineto $x $h $y add lineto closepath}def/concatprocs{/_p2
exch cvlit def/_p1 exch cvlit def/_pn _p1 length _p2 length add array def
_pn 0 _p1 putinterval _pn _p1 length _p2 putinterval _pn
cvx}def/OF/findfont load def/findfont{dup DEC_WRITE_dict exch
known{DEC_WRITE_dict exch get}if DEC_WRITE_dict/OF get exec}def
mark/ISOLatin1Encoding 
8#000 1 8#001{StandardEncoding exch get}for /emdash/endash
8#004 1 8#025{StandardEncoding exch get}for /quotedblleft/quotedblright
8#030 1 8#054{StandardEncoding exch get}for /minus 8#056 1 8#217
{StandardEncoding exch get}for/dotlessi 8#301 1 8#317{StandardEncoding 
exch get}for/space/exclamdown/cent/sterling/currency/yen/brokenbar/section
/dieresis/copyright/ordfeminine/guillemotleft/logicalnot/hyphen/registered
/macron/degree/plusminus/twosuperior/threesuperior/acute/mu/paragraph
/periodcentered/cedilla/onesuperior/ordmasculine/guillemotright/onequarter
/onehalf/threequarters/questiondown/Agrave/Aacute/Acircumflex/Atilde
/Adieresis/Aring/AE/Ccedilla/Egrave/Eacute/Ecircumflex/Edieresis/Igrave
/Iacute/Icircumflex/Idieresis/Eth/Ntilde/Ograve/Oacute/Ocircumflex/Otilde
/Odieresis/multiply/Oslash/Ugrave/Uacute/Ucircumflex/Udieresis/Yacute/Thorn
/germandbls/agrave/aacute/acircumflex/atilde/adieresis/aring/ae/ccedilla
/egrave/eacute/ecircumflex/edieresis/igrave/iacute/icircumflex/idieresis
/eth/ntilde/ograve/oacute/ocircumflex/otilde/odieresis/divide/oslash/ugrave
/uacute/ucircumflex/udieresis/yacute/thorn/ydieresis
256 array astore def cleartomark 
/encodefont{findfont dup maxlength dict begin{1 index/FID ne{def}{pop
pop}ifelse}forall/Encoding exch def dup/FontName exch def currentdict
definefont end}def/loads{/$/ISOLatin1Encoding load def/&/encodefont load
def/*/invertpattern load def/+/revertscreen load def/-/invertscreen load
def/:/concatprocs load def/^/setpattern load def/~/pat1 load def/_/pat2
load def/@/setrect load def/A/arcn load def/B/ashow load def/C/curveto load
def/D/def load def/E/eofill load def/F/findfont load def/G/setgray load
def/H/closepath load def/I/clip load def/K/kshow load def/L/lineto load
def/M/moveto load def/N/newpath load def/O/rotate load def/P/pop load
def/R/grestore load def/S/gsave load def/T/translate load def/U/sub load
def/V/div load def/W/widthshow load def/X/exch load def/Y/awidthshow load
def/a/save load def/c/setlinecap load def/d/setdash load def/e/restore load
def/f/setfont load def/g/initclip load def/h/show load def/i/setmiterlimit
load def/j/setlinejoin load def/k/stroke load def/l/rlineto load
def/m/rmoveto load def/n/currentfont load def/o/scalefont load
def/p/currentpoint load def/r/currenttransfer load def/s/scale load
def/t/setmatrix load def/u/settransfer load def/w/setlinewidth load
def/x/matrix load def/y/currentmatrix load def}def
end
%%EndProcSet
%%EndProlog

%%BeginSetup
DEC_WRITE_dict begin
loads
version cvi 23.0 gt {
currentdict {dup type /arraytype eq
{bind def} {pop pop} ifelse} forall} if
0.0100 0.0100 s

%%EndSetup
%%Page: 1 1
/$P a D
g N
0 79200 T
S
S
N
7500.00 -73600.00 M
53700.00 -73600.00 L
S
100 w
0 c
0 j
2 i
0.00 G k
R
R

R
S
7200 -7200 T
N
0 G
10664 -2100 M
/Helvetica-Bold-ISOLatin1 $
/Helvetica-Bold & P
/Helvetica-Bold-ISOLatin1 F 2400 o f
(European AI\255CC  FY91) h
7847 -5850 M
n 0.750 o f
(Conclusions and Recommendations) h
19083 -11150 M
/Times-Roman-ISOLatin1 $
/Times-Roman & P
/Times-Roman-ISOLatin1 F 1400 o f
(Joe  Fitzgibbon) h
13309 -16750 M
(Consultant,  EIC/ET\255IST  Valbonne) h
18966 -22200 M
n 0.857 o f
(22nd August 1991) h
300 -25800 M
300 -27800 M
300 -29800 M
300 -31800 M
300 -33800 M
300 -35800 M
300 -37800 M
300 -39800 M
300 -41800 M
300 -43800 M
300 -45800 M
300 -47800 M
300 -49800 M
300 -51800 M
300 -53800 M
300 -55800 M
300 -57800 M
300 -59800 M
-7200 7200 T
R

showpage
$P e

%%Page: 2 2
/$P a D
g N
0 79200 T
S
S
7200 -74700 T
N
0 G
22511 -1050 M
/Helvetica-ISOLatin1 $
/Helvetica & P
/Helvetica-ISOLatin1 F 1000 o f
(\255 ) h
(2) h
( \255) h
300 -2290 M
-7200 74700 T
R

S
N
7500.00 -73600.00 M
53700.00 -73600.00 L
S
100 w
0 c
0 j
2 i
0.00 G k
R
R

R
S
7200 -7200 T
N
0 G
300 -1350 M
/Helvetica-BoldOblique-ISOLatin1 $
/Helvetica-BoldOblique & P
/Helvetica-BoldOblique-ISOLatin1 F 1400 o f
(1  ) h
1857 -1350 M
(Management Summary.) h
300 -4200 M
6060 -4200 M
/Times-Roman-ISOLatin1 $
/Times-Roman & P
/Times-Roman-ISOLatin1 F 1200 o f
118.8 0 32 (The purpose of this report is to provide the Sales Support and Technology Group) W
300 -5600 M
(with a summary of,\255) h
300 -7600 M
/Symbol F 1200 o f
(\267) h
2100 -7600 M
/Times-Roman-ISOLatin1 F 1200 o f
(the recommendations supported by the European AI Competency Circle.) h
300 -9600 M
/Symbol F 1200 o f
(\267) h
2100 -9600 M
/Times-Roman-ISOLatin1 F 1200 o f
(the perceptions and consequent conclusions concerning the AI related business.) h
300 -11600 M
/Symbol F 1200 o f
(\267) h
2100 -11600 M
/Times-Roman-ISOLatin1 F 1200 o f
(the AI Competency Circle activities during FY91) h
300 -13600 M
/Symbol F 1200 o f
(\267) h
2100 -13600 M
/Times-Roman-ISOLatin1 F 1200 o f
(the AI Customer training program as implemented via the Fellowship Program.) h
300 -15600 M
/Symbol F 1200 o f
(\267) h
2100 -15600 M
/Times-Roman-ISOLatin1 F 1200 o f
12.1 0 32 (the AI related Customer project activity in three of the big countries, plus a view of the situ\255) W
2100 -17000 M
(ation in one of the small countries \(Belgium\).) h
300 -19000 M
300 -21950 M
/Helvetica-Bold-ISOLatin1 $
/Helvetica-Bold & P
/Helvetica-Bold-ISOLatin1 F 1400 o f
(1.1 ) h
2635 -21950 M
(Recommendations from the CC.) h
300 -24200 M
/Symbol F 1200 o f
(\267) h
2100 -24200 M
/Times-Roman-ISOLatin1 F 1200 o f
197.3 0 32 (Focus on those cross\255industry applications, and Consulting Services, which are likely to) W
2100 -25600 M
131.6 0 32 (benefit most from the integration of A.I. Techniques. These should then be referred to as) W
2100 -27000 M
135.4 0 32 ("Emerging Technologies" \(ET\), which is to be understood as the combining of A.I. tech\255) W
2100 -28400 M
(niques with other advanced technologies.  Some examples are.\255) h
2100 -30400 M
(\255) h
3900 -30400 M
(Intelligent Information Management and Retrieval. \( EXCALIBUR\)) h
2100 -32400 M
(\255) h
3900 -32400 M
(Multi\255media and Intelligent Tutoring applications. \(DIDDLY and MERLIN\)) h
2100 -34400 M
(\255) h
3900 -34400 M
221.8 0 32 (Strategic planning applications  \(e.g. KHARMA from Galway\) which use IMKA to) W
3900 -35800 M
(store, distribute and exercise the Customer's Knowledge assets.) h
2100 -37800 M
(\255) h
3900 -37800 M
(Dynamic Business Modeling.) h
2100 -39800 M
(\255) h
3900 -39800 M
23.8 0 32 (Intelligent Scheduling assistants, e.g. as applied to the Manufacturing and transportation) W
3900 -41200 M
(industries.) h
2100 -43200 M
(\255) h
3900 -43200 M
(Large scale configuration and diagnostic problems.) h
2100 -45200 M
(\255) h
3900 -45200 M
354.0 0 32 (Alarm correlation and decision support applied to the operational management of) W
3900 -46600 M
(Telecom/Computer networks, and computer systems.) h
300 -48600 M
/Symbol F 1200 o f
(\267) h
2100 -48600 M
/Times-Roman-ISOLatin1 F 1200 o f
146.0 0 32 (Establish an active Business development role for Emerging Technologies, \(ET\) with the) W
2100 -50000 M
(following objectives.) h
2100 -52000 M
(\255) h
3900 -52000 M
33.9 0 32 (Define a structured Portfolio of ET Services to be sold in Europe and identify European) W
3900 -53400 M
(level  Resources and Processes necessary for the support of that Portfolio.) h
2100 -55400 M
(\255) h
3900 -55400 M
13.3 0 32 (Work with Country teams to identify resources needed for the implementation of the ET) W
3900 -56800 M
(Services Portfolio.) h
2100 -58800 M
(\255) h
3900 -58800 M
164.4 0 32 (Co\255ordinate the activity of the European ET Competency Circle and represent needs,) W
3900 -60200 M
(opinions, to Marketing, Engineering, Educational Services, etc.) h
2100 -62200 M
(\255) h
3900 -62200 M
80.3 0 32 (Ensure the correct level of ET representation at the appropriate public trade shows and) W
3900 -63600 M
(Digital seminars or tradeshows \(e.g. DECville\).) h
-7200 7200 T
R

showpage
$P e

%%Page: 3 3
/$P a D
g N
0 79200 T
S
S
7200 -74700 T
N
0 G
22511 -1050 M
/Helvetica-ISOLatin1 $
/Helvetica & P
/Helvetica-ISOLatin1 F 1000 o f
(\255 ) h
(3) h
( \255) h
300 -2290 M
-7200 74700 T
R

S
N
7500.00 -73600.00 M
53700.00 -73600.00 L
S
100 w
0 c
0 j
2 i
0.00 G k
R
R

R
S
7200 -7200 T
N
0 G
300 -1350 M
/Helvetica-Bold-ISOLatin1 $
/Helvetica-Bold & P
/Helvetica-Bold-ISOLatin1 F 1400 o f
(1.2 ) h
2635 -1350 M
(Perceptions  and  Conclusions.) h
300 -4200 M
/Helvetica-BoldOblique-ISOLatin1 $
/Helvetica-BoldOblique & P
/Helvetica-BoldOblique-ISOLatin1 F 1200 o f
(1.2.1 ) h
3302 -4200 M
(AI Related Business Exists.) h
300 -6200 M
6000 -6200 M
/Times-Roman-ISOLatin1 $
/Times-Roman & P
/Times-Roman-ISOLatin1 F 1200 o f
157.6 0 32 (A recent Frost and Sullivan study, forecasts a total AI market for Europe of 382) W
300 -7600 M
27.7 0 32 (million Ecu in 1991, 520 million Ecu in  1992, and growing to 1286 million Ecu by 1995. \(Ad\255) W
300 -9000 M
(ditional detail can be found in paragraph 2 of this document.\)  ) h
300 -11600 M
6060 -11600 M
29.1 0 32 (AI related business exists; if direct Digital projects revenue is still too low, partners) W
300 -13000 M
70.0 0 32 (\(CSO's\) generated business is significant \(France makes heavy use of partners due to shortage) W
300 -14400 M
138.7 0 32 (of internal expertise and their relatively low cost per man\); projects where Digital is directly) W
300 -15800 M
0.6 0 32 (implied are about to generate important leverage business \(account penetration \255 e.g. the BNP in) W
300 -17200 M
16.0 0 32 (France and GD in Italy which was an IBM stronghold\) which is unfortunately difficult to meas\255) W
300 -18600 M
(ure.) h
300 -20600 M
6060 -20600 M
134.4 0 32 (As an example of the customer interest in this space, the Emerging Technologies) W
300 -22000 M
50.4 0 32 (Customer Seminar, which is to be held at the Valbonne Executive Customer Centre in October) W
300 -23400 M
56.1 0 32 (1991, is currently oversubscribed with more than 50 customers having made reservations. This) W
300 -24800 M
(seminar is hosted by the ET\255IST, part of the Valbonne EIC.) h
300 -26800 M
6060 -26800 M
131.5 0 32 (Given that this business exists, a \(better co\255ordinated\) Marketing effort which es\255) W
300 -28200 M
141.1 0 32 (tablishes a clear view for the Account Teams of Digital's expertise, products and services in) W
300 -29600 M
(this space, would support the generation of significant additional revenues.) h
300 -31600 M
300 -34200 M
/Helvetica-BoldOblique-ISOLatin1 F 1200 o f
(1.2.2 ) h
3302 -34200 M
( Focus on Strategic, Cross\255industry, Applications.) h
300 -36200 M
6060 -36200 M
/Times-Roman-ISOLatin1 F 1200 o f
56.7 0 32 (The acronym AI is mis\255leading and causes confusion. We have eliminated this dif\255) W
300 -37600 M
24.1 0 32 (ficulty by talking of more tangible examples of the use of A.I. software technology as  "Emerg\255) W
300 -39000 M
82.4 0 32 (ing Technology" \(ET\). Under the umbrella of ET we currently refer to base technologies such) W
300 -40400 M
19.7 0 32 (as "Object Oriented", "Neural Networks", "Constraint Logic Programming", or to classes of ap\255) W
300 -41800 M
243.1 0 32 (plications such as "Multi\255media", "Voice Interfaces", "Scheduling", "Intelligent Information) W
300 -43200 M
(Management", "Intelligent Tutoring Systems".) h
300 -45200 M
6060 -45200 M
109.9 0 32 (Each of these terms are intended to allow the field operative, who is faced with a) W
300 -46600 M
109.8 0 32 (customer problem, to more easily associate a possible solution source with that problem. It is) W
300 -48000 M
90.9 0 32 (worth noting that each of the above application classes identify some of the strategic business) W
300 -49400 M
(problems facing our customers today, and they are Cross\255industry.) h
300 -51400 M
300 -54000 M
/Helvetica-BoldOblique-ISOLatin1 F 1200 o f
(1.2.3 ) h
3302 -54000 M
(Speed up Revenue Generation.) h
300 -56000 M
6060 -56000 M
/Times-Roman-ISOLatin1 F 1200 o f
73.8 0 32 (The rate at which a project generates revenue, compared to resources consumed in) W
300 -57400 M
102.7 0 32 (providing the solution, can be better balanced \(i.e. in Digital's favour\) by the provision of re\255) W
300 -58800 M
117.2 0 32 (usable software solutions platforms. Examples of this are EXCALIBUR, DIDDLY, MERLIN) W
300 -60200 M
(ATEM, UFIS and SCHEDULER.) h
300 -62200 M
-7200 7200 T
R

showpage
$P e

%%Page: 4 4
/$P a D
g N
0 79200 T
S
S
7200 -74700 T
N
0 G
22511 -1050 M
/Helvetica-ISOLatin1 $
/Helvetica & P
/Helvetica-ISOLatin1 F 1000 o f
(\255 ) h
(4) h
( \255) h
300 -2290 M
-7200 74700 T
R

S
N
7500.00 -73600.00 M
53700.00 -73600.00 L
S
100 w
0 c
0 j
2 i
0.00 G k
R
R

R
S
7200 -7200 T
N
0 G
300 -1200 M
/Helvetica-BoldOblique-ISOLatin1 $
/Helvetica-BoldOblique & P
/Helvetica-BoldOblique-ISOLatin1 F 1200 o f
(1.2.4 ) h
3302 -1200 M
(There are too few AI Specialists.) h
300 -3200 M
6060 -3200 M
/Times-Roman-ISOLatin1 $
/Times-Roman & P
/Times-Roman-ISOLatin1 F 1200 o f
37.4 0 32 (There are too few AI Specialists to manage on\255going projects and lengthy pre\255sales) W
300 -4600 M
39.3 0 32 (activities: much of the AI business is lost to partners or competition, e.g. Digital lost  the SOL\255) W
300 -6000 M
(LAC's SACHEM 80MFF \(approx  14M$\) Black Furnace project to a BULL/Cap\255Sesa team.) h
300 -8000 M
6060 -8000 M
80.0 0 32 (Countries should be encouraged to develop Specialists with some awareness of AI) W
300 -9400 M
(methods and tools, not necessarily fully trained  Knowledge Engineers.) h
300 -11400 M
300 -14000 M
/Helvetica-BoldOblique-ISOLatin1 F 1200 o f
(1.2.5 ) h
3302 -14000 M
(A \(better co\255ordinated\) Marketing role.) h
300 -16000 M
6060 -16000 M
/Times-Roman-ISOLatin1 F 1200 o f
169.7 0 32 (When asking customers why they did not come to Digital for a solution to their) W
300 -17400 M
94.7 0 32 (\(A.I.\) problem, the response is often one of surprise to learn that we can offer services in this) W
300 -18800 M
94.0 0 32 (space. i.e. We are failing to adequately advertise our capabilities to both customers and to ac\255) W
300 -20200 M
(count teams.) h
300 -22200 M
6060 -22200 M
98.6 0 32 (A \(better co\255ordinated\) Marketing role in the definition and  selling of  AI related) W
300 -23600 M
3.2 0 32 (services, not just software and hardware products, would benefit the country level teams signifi\255) W
300 -25000 M
52.8 0 32 (cantly. This role would ensure that the Account Managers are aware of \(and know how to use\)) W
300 -26400 M
31.4 0 32 (the ET opportunities and the available support services, as well as identifying the factors which) W
300 -27800 M
(differentiate the Digital offering from that of the competition.) h
300 -29800 M
300 -32400 M
/Helvetica-BoldOblique-ISOLatin1 F 1200 o f
(1.2.6 ) h
3302 -32400 M
(Training mis\255match.) h
300 -34400 M
6060 -34400 M
/Times-Roman-ISOLatin1 F 1200 o f
39.8 0 32 (Projects often include demands for training, demands which did not match with the) W
300 -35800 M
39.8 0 32 (Ed. services catalogue of training offers which were viewed as being inadequate. Consequently) W
300 -37200 M
19.3 0 32 (training has been designed and given by country AI Specialists. There may be a role for ET\255CC) W
300 -38600 M
8.3 0 32 (members which would \(partially\) resolve this difficulty. However, given the wide variety of 3rd) W
300 -40000 M
128.8 0 32 (party tools which are available on Digital hardware, under DDS, CMP and CSH agreements,) W
300 -41400 M
111.5 0 32 (some clarity of availability of training for these tools is required. e.g. The Corporate AITC is) W
300 -42800 M
210.8 0 32 (offering, to internal and external customers, training on Tools \(both Digital and 3rd party\),) W
300 -44200 M
(Methodologies, and on the Management of A.I. based projects.  ) h
300 -46200 M
300 -48200 M
300 -50200 M
300 -52200 M
300 -54200 M
300 -56200 M
300 -58200 M
300 -60200 M
-7200 7200 T
R

showpage
$P e

%%Page: 5 5
/$P a D
g N
0 79200 T
S
S
N
7500.00 -73600.00 M
53700.00 -73600.00 L
S
100 w
0 c
0 j
2 i
0.00 G k
R
R

R
S
7200 -7200 T
N
0 G
300 -1350 M
/Helvetica-Bold-ISOLatin1 $
/Helvetica-Bold & P
/Helvetica-Bold-ISOLatin1 F 1400 o f
(1.3 ) h
2635 -1350 M
(Summary of AI Competency Circle Activity 1991.) h
300 -3600 M
6060 -3600 M
/Times-Roman-ISOLatin1 $
/Times-Roman & P
/Times-Roman-ISOLatin1 F 1200 o f
148.6 0 32 (Only the first of the planned two meetings were held this year, the second being) W
300 -5000 M
(cancelled due to travel restrictions imposed for the duration of the Gulf war.) h
300 -7000 M
300 -9600 M
/Helvetica-BoldOblique-ISOLatin1 $
/Helvetica-BoldOblique & P
/Helvetica-BoldOblique-ISOLatin1 F 1200 o f
(1.3.1 ) h
3302 -9600 M
(Meeting of 29\25531 October 1990.) h
300 -11600 M
6060 -11600 M
/Times-Roman-ISOLatin1 F 1200 o f
88.2 0 32 (The first meeting was held the 29\25531 October 1990 during the week following the) W
300 -13000 M
31.0 0 32 (Emerging Technologies Customer Seminar, which was hosted at Valbonne Executive centre by) W
300 -14400 M
(the  ET\255IST \(EIC\). The following conclusions were drawn from this meeting.) h
300 -16400 M
300 -18400 M
/Symbol F 1200 o f
(\267) h
2100 -18400 M
/Times-Roman-ISOLatin1 F 1200 o f
80.0 0 32 (The direction established by the ET\255IST in focusing on the practical application aspects of) W
2100 -19800 M
106.7 0 32 (AI under the banner of  "Emerging Technologies" was generally  agreed to be the correct) W
2100 -21200 M
(approach.) h
300 -23200 M
/Symbol F 1200 o f
(\267) h
2100 -23200 M
/Times-Roman-ISOLatin1 F 1200 o f
95.4 0 32 (We need to focus available resources on applications which offer a competitive advantage) W
2100 -24600 M
100.5 0 32 (through the integration of A.I. techniques with other Emerging Technologies; these would) W
2100 -26000 M
4.8 0 32 (ideally be cross\255industry applications, e.g. Intelligent Information Management, Scheduling,) W
2100 -27400 M
(etc.) h
300 -29400 M
/Symbol F 1200 o f
(\267) h
2100 -29400 M
/Times-Roman-ISOLatin1 F 1200 o f
117.4 0 32 (In order to be more competitive, and with a view to achieving shorter revenue generation) W
2100 -30800 M
119.5 0 32 (time frames, we should be prepared to offer more in  the way of re\255usable software solu\255) W
2100 -32200 M
(tions platforms e.g. EXCALIBUR, DIDDLY, ATEM, UFIS, MERLIN or SCHEDULER.) h
300 -34200 M
/Symbol F 1200 o f
(\267) h
2100 -34200 M
/Times-Roman-ISOLatin1 F 1200 o f
97.5 0 32 (Corporate AI marketing's strategy in favouring CSH product agreements over CMP, DDS) W
2100 -35600 M
140.6 0 32 (or Digital products results in  shortage of Digital products on RISC platforms, and could) W
2100 -37000 M
(lead to different AI Product Portfolios for each country.) h
300 -39000 M
300 -41000 M
6060 -41000 M
1.1 0 32 (A report on this meeting, produced by Dirk Bade of Germany, can be found in para\255) W
300 -42400 M
(graph 2 of this document.) h
300 -44400 M
300 -46400 M
300 -48400 M
300 -50400 M
300 -52400 M
300 -54400 M
300 -56400 M
300 -58400 M
300 -60400 M
300 -62400 M
-7200 7200 T
R

showpage
$P e

%%Page: 6 6
/$P a D
g N
0 79200 T
S
S
7200 -74700 T
N
0 G
22511 -1050 M
/Helvetica-ISOLatin1 $
/Helvetica & P
/Helvetica-ISOLatin1 F 1000 o f
(\255 ) h
(6) h
( \255) h
300 -2290 M
-7200 74700 T
R

S
N
7500.00 -73600.00 M
53700.00 -73600.00 L
S
100 w
0 c
0 j
2 i
0.00 G k
R
R

R
S
7200 -7200 T
N
0 G
300 -1350 M
/Helvetica-Bold-ISOLatin1 $
/Helvetica-Bold & P
/Helvetica-Bold-ISOLatin1 F 1400 o f
(1.4 ) h
2635 -1350 M
(European AI Fellowship.) h
300 -3600 M
6060 -3600 M
/Times-Roman-ISOLatin1 $
/Times-Roman & P
/Times-Roman-ISOLatin1 F 1200 o f
181.7 0 32 (This is a Corporate Program financed from strategic funds and designed to con\255) W
300 -5000 M
46.6 0 32 (vince important customers of Digitals leadership in the field of practical AI applications. FY91) W
300 -6400 M
3.8 0 32 (is the first year in which this program has been held at a location outside of the US based AITC,) W
300 -7800 M
31.4 0 32 (as such it represents an important investment on the part of Corporate. The following European) W
300 -9200 M
(Customers sent a student to the FY91 AI Fellowship offering, which was held in Valbonne.) h
300 -11200 M
/Symbol F 1200 o f
(\267) h
2100 -11200 M
/Times-Roman-ISOLatin1 F 1200 o f
(British Telecomm.) h
300 -13200 M
/Symbol F 1200 o f
(\267) h
2100 -13200 M
/Times-Roman-ISOLatin1 F 1200 o f
(Phillips of Holland.) h
300 -15200 M
/Symbol F 1200 o f
(\267) h
2100 -15200 M
/Times-Roman-ISOLatin1 F 1200 o f
(ENEA of Italy.) h
300 -17200 M
/Symbol F 1200 o f
(\267) h
2100 -17200 M
/Times-Roman-ISOLatin1 F 1200 o f
(Irish Telecomm.) h
300 -19200 M
6060 -19200 M
188.2 0 32 (Due to European constraints of timescale and resources, the following European) W
300 -20600 M
(Customers sent students to the Corporate program during FY91.) h
300 -22600 M
/Symbol F 1200 o f
(\267) h
2100 -22600 M
/Times-Roman-ISOLatin1 F 1200 o f
(Volkswagen \(Germany\)) h
300 -24600 M
/Symbol F 1200 o f
(\267) h
2100 -24600 M
/Times-Roman-ISOLatin1 F 1200 o f
(Dutch PTT) h
300 -26600 M
/Symbol F 1200 o f
(\267) h
2100 -26600 M
/Times-Roman-ISOLatin1 F 1200 o f
(CIT\255ALCATEL \(France\)) h
300 -28600 M
/Symbol F 1200 o f
(\267) h
2100 -28600 M
/Times-Roman-ISOLatin1 F 1200 o f
(Mercury Communications \(UK\)) h
300 -30600 M
/Symbol F 1200 o f
(\267) h
2100 -30600 M
/Times-Roman-ISOLatin1 F 1200 o f
(Telespatio \(Italy\)) h
300 -32600 M
/Symbol F 1200 o f
(\267) h
2100 -32600 M
/Times-Roman-ISOLatin1 F 1200 o f
(SMA \(Italy\)) h
300 -34600 M
300 -36600 M
6060 -36600 M
15.0 0 32 (The FY92 AI Fellowship Program activity will be paid for by the customer, at 48k$) W
300 -38000 M
109.2 0 32 (per head, with 5\2556 full\255time students scheduled for the Q2 session. In addition the FY92 Pro\255) W
300 -39400 M
84.2 0 32 (gram content has been significantly changed subsequent to feedback received from the AI\255CC) W
300 -40800 M
(on European Customer expectations.) h
300 -42800 M
6060 -42800 M
31.8 0 32 (The detailed report on the results of the first European AI Fellowship Program held) W
300 -44200 M
(in Valbonne, January \255 June 1991, can be found in para. 2 of this report.) h
300 -46200 M
300 -48200 M
300 -50200 M
300 -52200 M
300 -54200 M
300 -56200 M
300 -58200 M
-7200 7200 T
R

showpage
$P e

%%Page: 7 7
/$P a D
g N
0 79200 T
S
S
7200 -74700 T
N
0 G
22511 -1050 M
/Helvetica-ISOLatin1 $
/Helvetica & P
/Helvetica-ISOLatin1 F 1000 o f
(\255 ) h
(7) h
( \255) h
300 -2290 M
-7200 74700 T
R

S
N
7500.00 -73600.00 M
53700.00 -73600.00 L
S
100 w
0 c
0 j
2 i
0.00 G k
R
R

R
S
7200 -7200 T
N
0 G
300 -1350 M
/Helvetica-Bold-ISOLatin1 $
/Helvetica-Bold & P
/Helvetica-Bold-ISOLatin1 F 1400 o f
(1.5 ) h
2635 -1350 M
( Customer A.I. Project Activity During FY91.) h
300 -4200 M
/Helvetica-BoldOblique-ISOLatin1 $
/Helvetica-BoldOblique & P
/Helvetica-BoldOblique-ISOLatin1 F 1200 o f
(1.5.1 ) h
3302 -4200 M
(France \(Stephane Ripoll\).) h
300 -6200 M
/Symbol F 1200 o f
(\267) h
2100 -6200 M
/Times-Roman-ISOLatin1 $
/Times-Roman & P
/Times-Roman-ISOLatin1 F 1200 o f
(Business  status for projects for which we have a direct visibility :) h
2100 -8200 M
(\255) h
3900 -8200 M
(direct Digital projects : 1.5 MFF ;) h
2100 -10200 M
(\255) h
3900 -10200 M
(close 3rd parties : more than 2.0 MFF s/w ) h
2100 -12200 M
(\255) h
3900 -12200 M
171.9 0 32 (many projects can't be done/are lost due to late implication in the selling cycle \(e.g.) W
3900 -13600 M
(Schlumberger, Soretex, Airbus Consortium Order Mgt system, ...\) ;) h
2100 -15600 M
(\255) h
3900 -15600 M
(projects made by Digital France often subcontracted ;) h
2100 -17600 M
(\255) h
3900 -17600 M
(many projects "left" to 3rd parties without control, for lack\255of\255resources reasons.) h
3900 -19000 M
300 -21000 M
/Symbol F 1200 o f
(\267) h
2100 -21000 M
/Times-Roman-ISOLatin1 F 1200 o f
(A.I. Projects.) h
2100 -23000 M
(\255) h
3900 -23000 M
(CIT\255ALCATEL. AI Fellowship, Telephone Network Monitoring expert system,) h
2100 -25000 M
(\255) h
3900 -25000 M
244.2 0 32 (COM'AZUR. Tourist enquiry expert server offering vocie and Minitel interface via) W
3900 -26400 M
(public telephone network \(in French and English\).) h
2100 -28400 M
(\255) h
3900 -28400 M
(ROQUETTE.  Planning and scheduling expert system for the food industry.) h
2100 -30400 M
(\255) h
3900 -30400 M
(VALLOUREC. BOM generation expert system, prototype presented in May.) h
2100 -32400 M
(\255) h
3900 -32400 M
(SOCS. Conversion of Banking application from VM/PROLOG to PROLOG. ) h
2100 -34400 M
(\255) h
3900 -34400 M
(ASF. Autoroute toll booth staff scheduling expert system.) h
2100 -36400 M
(\255) h
3900 -36400 M
(BNP, study \255 problem of application software configuration and testing.) h
3900 -37800 M
300 -39800 M
/Symbol F 1200 o f
(\267) h
2100 -39800 M
/Times-Roman-ISOLatin1 F 1200 o f
(Pre\255sales activity .) h
2100 -41800 M
(\255) h
3900 -41800 M
44.2 0 32 (COGEMA : screening of computer user calls using expert systems and voice interface ;) W
3900 -43200 M
(process diagnosis and monitoring to be discussed ;) h
2100 -45200 M
(\255) h
3900 -45200 M
( ISL : AI Fellowship to start by the end of calendar year ;  ) h
2100 -47200 M
(\255) h
3900 -47200 M
263.6 0 32 (Lille Hospital : toxicologic problems monitoring system ; proposal for a feasibility) W
3900 -48600 M
(study ;) h
2100 -50600 M
(\255) h
3900 -50600 M
(DCN : expert systems feasibility proposal for managing Technical Instructions ) h
2100 -52600 M
(\255) h
3900 -52600 M
(PROST : interest for AI and transport ;) h
2100 -54600 M
(\255) h
3900 -54600 M
(Peugeot : meeting planned with Technology Mgmt.) h
2100 -56600 M
(\255) h
3900 -56600 M
153.2 0 32 (3rd parties selling projects for Aeroports de Paris and Martell, Cr\351dit Agricole, CCF,) W
3900 -58000 M
(Cr\351dit Lyonnais...) h
300 -60000 M
-7200 7200 T
R

showpage
$P e

%%Page: 8 8
/$P a D
g N
0 79200 T
S
S
7200 -74700 T
N
0 G
22511 -1050 M
/Helvetica-ISOLatin1 $
/Helvetica & P
/Helvetica-ISOLatin1 F 1000 o f
(\255 ) h
(8) h
( \255) h
300 -2290 M
-7200 74700 T
R

S
N
7500.00 -73600.00 M
53700.00 -73600.00 L
S
100 w
0 c
0 j
2 i
0.00 G k
R
R

R
S
7200 -7200 T
N
0 G
300 -1200 M
/Helvetica-BoldOblique-ISOLatin1 $
/Helvetica-BoldOblique & P
/Helvetica-BoldOblique-ISOLatin1 F 1200 o f
(1.5.2 ) h
3302 -1200 M
(Italy \(Riccardo Simino\).) h
300 -3200 M
/Times-Roman-ISOLatin1 $
/Times-Roman & P
/Times-Roman-ISOLatin1 F 1200 o f
(  ) h
300 -5800 M
/Symbol F 1200 o f
(\267) h
2100 -5800 M
/Times-Roman-ISOLatin1 F 1200 o f
120.7 0 32 ( ENEA . The IDEA Project: Use of Neural Nets, AI, OO Technology for three important) W
2100 -7200 M
300.9 0 32 (research streams: Robotics and Vision; Advanced information management and Office) W
2100 -8600 M
(Automation; Environment management.) h
2100 -11200 M
(\255) h
3900 -11200 M
49.2 0 32 (LEVERAGED HW: VAX 9440 \(largest ever sold by the Corporation\) that will be used) W
3900 -12600 M
(for the development of the projects) h
2100 -15200 M
(\255) h
3900 -15200 M
70.9 0 32 (LEVERAGED CONSULTANCY: 36 MAN/YEARS for the development of these pro\255) W
3900 -16600 M
(jects.) h
300 -19200 M
/Symbol F 1200 o f
(\267) h
2100 -19200 M
/Times-Roman-ISOLatin1 F 1200 o f
18.7 0 32 (BONIFICA. RASVECTOR Project: Using Neural Nets and Object Oriented database to ex\255) W
2100 -20600 M
(tract manage and save information from satellite photos for mapping.) h
2100 -23200 M
(\255) h
3900 -23200 M
(LEVERAGED HW: 2 * VAX9410 \(One devoted to the project development\)) h
300 -25800 M
/Symbol F 1200 o f
(\267) h
2100 -25800 M
/Times-Roman-ISOLatin1 F 1200 o f
8.8 0 32 ( DEFENSE MINISTRY: Top Secret Project using Neural Nets. HIGHLY  STRATEGIC for) W
2100 -27200 M
(Digital Italy.) h
300 -29800 M
/Symbol F 1200 o f
(\267) h
2100 -29800 M
/Times-Roman-ISOLatin1 F 1200 o f
(ITALCABLE: Two Project concerning the use of Neural nets to develop an  ) h
40200 -29800 M
(information) h
2100 -31200 M
(management system for their technical manuals) h
2100 -33800 M
(\255) h
3900 -33800 M
(LEVERAGED CONSULTANCY about 200K$ ) h
300 -36400 M
/Symbol F 1200 o f
(\267) h
2100 -36400 M
/Times-Roman-ISOLatin1 F 1200 o f
105.9 0 32 (PIRELLI \255 WHIRLPOOL: Expert system routinely used by Whirlpool for the  production) W
2100 -37800 M
(scheduling of European frigiders.) h
300 -40400 M
/Symbol F 1200 o f
(\267) h
2100 -40400 M
/Times-Roman-ISOLatin1 F 1200 o f
(MARS. AJUTOR project.: an intelligent tutoring application for the field of micro\255gravity ) h
300 -43000 M
/Symbol F 1200 o f
(\267) h
2100 -43000 M
/Times-Roman-ISOLatin1 F 1200 o f
( GD:   ) h
6000 -43000 M
41.0 0 32 (DAYSY Project: AI has helped us in cracking in an IBM shop. Actually, after hav\255) W
2100 -44400 M
93.5 0 32 (ing sold 20 workstations, some PDP and servers,  Digital, shown to be the most advanced) W
2100 -45800 M
(technology partner using AI,  is going to sell a 9000.   ) h
300 -48400 M
/Symbol F 1200 o f
(\267) h
2100 -48400 M
/Times-Roman-ISOLatin1 F 1200 o f
136.2 0 32 (POLICE: Two starting projects: The first is TOP SECRET and cannot be described \(will) W
2100 -49800 M
23.9 0 32 (use Neural Nets\) but is considered highly strategic. The second is a training system   for the) W
2100 -51200 M
(Police School \(probably based on Merlin\)   ) h
300 -53800 M
/Symbol F 1200 o f
(\267) h
2100 -53800 M
/Times-Roman-ISOLatin1 F 1200 o f
( FINANCE POLICE:  Starting Project to develop an Expert System to help policemen in) h
2100 -55200 M
80.9 0 32 ( control if the Companies avoid to pay the taxes.  I would to stress that all the projects in\255) W
2100 -56600 M
47.3 0 32 (volving Defense Ministry and  Police are for us very important because are a powerful tool) W
2100 -58000 M
(to go into  the power centers of the Government.   ) h
300 -60600 M
/Symbol F 1200 o f
(\267) h
2100 -60600 M
/Times-Roman-ISOLatin1 F 1200 o f
105.3 0 32 ( ITALIAN BANK \(CRT Turin\): Two projects using Expert systems \(one ended  and rou\255) W
2100 -62000 M
(tinely used\) for the Credit analysis.   ) h
-7200 7200 T
R

showpage
$P e

%%Page: 9 9
/$P a D
g N
0 79200 T
S
S
7200 -74700 T
N
0 G
22511 -1050 M
/Helvetica-ISOLatin1 $
/Helvetica & P
/Helvetica-ISOLatin1 F 1000 o f
(\255 ) h
(9) h
( \255) h
300 -2290 M
-7200 74700 T
R

S
N
7500.00 -73600.00 M
53700.00 -73600.00 L
S
100 w
0 c
0 j
2 i
0.00 G k
R
R

R
S
7200 -7200 T
N
0 G
300 -1200 M
/Symbol F 1200 o f
(\267) h
2100 -1200 M
/Times-Roman-ISOLatin1 $
/Times-Roman & P
/Times-Roman-ISOLatin1 F 1200 o f
109.3 0 32 ( ALFA AVIO: Starting project to develop an Expert System for failure  repairing of heli\255) W
2100 -2600 M
13.9 0 32 (copters engines. This project for us is very strategic  to leverage the business concerning the) W
2100 -4000 M
(realization of the second  Factory of Alfa Romeo Avionics.) h
2100 -6600 M
(\255) h
3900 -6600 M
(business leveraged about 4 Million $) h
300 -9200 M
(.    ) h
300 -12400 M
/Helvetica-BoldOblique-ISOLatin1 $
/Helvetica-BoldOblique & P
/Helvetica-BoldOblique-ISOLatin1 F 1200 o f
(1.5.3 ) h
3302 -12400 M
(Belgium \(Luc Vandevyver\)) h
300 -14400 M
/Symbol F 1200 o f
(\267) h
2100 -14400 M
/Times-Roman-ISOLatin1 F 1200 o f
41.7 0 32 (We \(digital Belgium\) have not realised any customer project with AI\255contents yet. The rea\255) W
2100 -15800 M
(sons for this :) h
2100 -18400 M
(\255) h
3900 -18400 M
(limited local support \(1.5 people\)) h
2100 -21000 M
(\255) h
3900 -21000 M
(distance to Valbonne) h
2100 -23600 M
(\255) h
3900 -23600 M
146.1 0 32 (AI techniques helped to better understand problem, which eventually could be solved) W
3900 -25000 M
(the traditional way.) h
300 -27600 M
/Symbol F 1200 o f
(\267) h
2100 -27600 M
/Times-Roman-ISOLatin1 F 1200 o f
210.8 0 32 (But we do have customers which are active or considering using the technology. Some) W
2100 -29000 M
(names:) h
2100 -31600 M
(\255) h
3900 -31600 M
( SIDMAR \(steel factory\); continuous research for better production methods) h
2100 -34200 M
(\255) h
3900 -34200 M
(COCKERILL \(steel factory\); process monitoring) h
2100 -36800 M
(\255) h
3900 -36800 M
(Forges de Glabecq \(steel factory\);  production planning) h
2100 -39400 M
(\255) h
3900 -39400 M
( Picanol \(machine builders\); fault diagnosis) h
2100 -42000 M
(\255) h
3900 -42000 M
( Phillips; planning) h
2100 -44600 M
(\255) h
3900 -44600 M
( BIKIT \(University research\); several customer projects) h
2100 -47200 M
(\255) h
3900 -47200 M
( SWIFT \(banking network\); continuous research) h
300 -49800 M
/Symbol F 1200 o f
(\267) h
2100 -49800 M
/Times-Roman-ISOLatin1 F 1200 o f
(Besides the customers, some Solution Providers are very active in the Belgian market :) h
2100 -52400 M
(\255) h
3900 -52400 M
( AICORP \(KBMS\)) h
2100 -55000 M
(\255) h
3900 -55000 M
( GENSYM \(G2\)) h
2100 -57600 M
(\255) h
3900 -57600 M
( AI\255SYSTEMS \(G2, KBMS, Nexpert, projects\)) h
2100 -60200 M
(\255) h
3900 -60200 M
(Software Generation \(ADS\)) h
2100 -62800 M
(\255) h
3900 -62800 M
( E2S \(projects\)) h
-7200 7200 T
R

showpage
$P e

%%Page: 10 10
/$P a D
g N
0 79200 T
S
S
7200 -74700 T
N
0 G
22233 -1050 M
/Helvetica-ISOLatin1 $
/Helvetica & P
/Helvetica-ISOLatin1 F 1000 o f
(\255 ) h
(10) h
( \255) h
300 -2290 M
-7200 74700 T
R

S
N
7500.00 -73600.00 M
53700.00 -73600.00 L
S
100 w
0 c
0 j
2 i
0.00 G k
R
R

R
S
7200 -7200 T
N
0 G
300 -1200 M
/Helvetica-BoldOblique-ISOLatin1 $
/Helvetica-BoldOblique & P
/Helvetica-BoldOblique-ISOLatin1 F 1200 o f
(1.5.4 ) h
3302 -1200 M
(Germany \(Dirk Bade\).) h
300 -3200 M
6000 -3200 M
/Times-Roman-ISOLatin1 $
/Times-Roman & P
/Times-Roman-ISOLatin1 F 1200 o f
(In GY our AI\255activities in can be summarized as follows:) h
300 -5800 M
/Symbol F 1200 o f
(\267) h
2100 -5800 M
/Times-Roman-ISOLatin1 F 1200 o f
141.8 0 32 (We concentrate on marketing two major efforts \(UFIS and ATEM\) and try push forward) W
2100 -7200 M
(associative memory/neural net technology. For ATEM we are driving projects with:) h
2100 -9800 M
(\255) h
3900 -9800 M
(Philips Medical Systems \(Hamburg\) will receive test installation   ) h
2100 -12400 M
(\255) h
3900 -12400 M
(Deutsche Airbus GmbH \(Hamburg\) received test installation  ) h
2100 -15000 M
(\255) h
3900 -15000 M
(BMW \(M\374nchen\) may wish to integrate ATEM in own system) h
2100 -17600 M
(\255) h
3900 -17600 M
( Lufthansa \(Frankfurt\) received DEMO) h
2100 -20200 M
(\255) h
3900 -20200 M
( Dupont \(Frankfurt\) received test installation) h
2100 -22800 M
(\255) h
3900 -22800 M
( debis \(Stuttgart\) received DEMO) h
2100 -25400 M
(\255) h
3900 -25400 M
( Swiss PTT \(Bern\), now locally driven) h
2100 -28000 M
(\255) h
3900 -28000 M
( Stadt Wien \(Vienna\), will be locally driven) h
300 -30600 M
/Symbol F 1200 o f
(\267) h
2100 -30600 M
/Times-Roman-ISOLatin1 F 1200 o f
28.9 0 32 (A presentation on "Distributed Multimedia Applications using DECwindows" was given on) W
2100 -32000 M
264.5 0 32 (the ONLINE fair in Hamburg and essentially was on ATEM. A Market Requirements) W
2100 -33400 M
218.6 0 32 (Document on ATEM was composed and submitted for  productisation. DACH network) W
2100 -34800 M
(marketing now wants ATEM\255like additions  to DECmcc.) h
300 -37400 M
/Symbol F 1200 o f
(\267) h
2100 -37400 M
/Times-Roman-ISOLatin1 F 1200 o f
133.3 0 32 (UFIS is driven locally in Stuttgart by Karl Kramer @SUT, who works on the phase 2 of) W
2100 -38800 M
114.1 0 32 (UFIS and on country level by Manfred Schlueter @SUT, who is busy with acquisition of) W
2100 -40200 M
(UFIS\255like follow\255on projects: ) h
2100 -42800 M
(\255) h
3900 -42800 M
( McKinsey,  ) h
2100 -45400 M
(\255) h
3900 -45400 M
(BASF, ) h
2100 -48000 M
(\255) h
3900 -48000 M
( SEL,  ) h
2100 -50600 M
(\255) h
3900 -50600 M
( Bayer   and others) h
300 -53200 M
/Symbol F 1200 o f
(\267) h
2100 -53200 M
/Times-Roman-ISOLatin1 F 1200 o f
49.5 0 32 (UFIS was shown very successful on the Hannover CEBIT fair. Please mail  Karl and Man\255) W
2100 -54600 M
(fred if you need more details.) h
300 -57200 M
/Symbol F 1200 o f
(\267) h
2100 -57200 M
/Times-Roman-ISOLatin1 F 1200 o f
(Peter Brittain from our ACT supports fulltime the Swiss SANDOZ Project.) h
300 -59800 M
/Symbol F 1200 o f
(\267) h
2100 -59800 M
/Times-Roman-ISOLatin1 F 1200 o f
9.6 0 32 (Presales support on configuration problems was given to PKI\255N\374rnberg and two NEXPERT) W
2100 -61200 M
(OBJECT licenses will probably be the next sale.) h
300 -63800 M
/Symbol F 1200 o f
(\267) h
2100 -63800 M
/Times-Roman-ISOLatin1 F 1200 o f
21.9 0 32 (Associative Memory \(PAM\) and Neural Net Application \(EXCALIBUR\) was presented  on) W
-7200 7200 T
R

showpage
$P e

%%Page: 11 11
/$P a D
g N
0 79200 T
S
S
7200 -74700 T
N
0 G
22233 -1050 M
/Helvetica-ISOLatin1 $
/Helvetica & P
/Helvetica-ISOLatin1 F 1000 o f
(\255 ) h
(11) h
( \255) h
300 -2290 M
-7200 74700 T
R

S
N
7500.00 -73600.00 M
53700.00 -73600.00 L
S
100 w
0 c
0 j
2 i
0.00 G k
R
R

R
S
7200 -7200 T
N
0 G
2100 -1200 M
/Times-Roman-ISOLatin1 $
/Times-Roman & P
/Times-Roman-ISOLatin1 F 1200 o f
(a Customer Seminar to customers from WEGY:) h
2100 -3800 M
(\255) h
3900 -3800 M
(Schlumberger, ) h
2100 -6400 M
(\255) h
3900 -6400 M
(Flachglas AG) h
2100 -9000 M
(\255) h
3900 -9000 M
(Spinnbaum AG) h
2100 -11600 M
(\255) h
3900 -11600 M
( Mannesmann) h
300 -14200 M
/Symbol F 1200 o f
(\267) h
2100 -14200 M
/Times-Roman-ISOLatin1 F 1200 o f
133.8 0 32 (In addition, we have supported CASE marketing on the German Riverboat Show and the) W
2100 -15600 M
(Hannover CEBIT fair in preparing and presenting a user interface tool \(VUIT\) demo. ) h
300 -18200 M
300 -20800 M
300 -23400 M
300 -26000 M
300 -28600 M
300 -31200 M
300 -33800 M
300 -36400 M
300 -39000 M
300 -41600 M
300 -44200 M
300 -46800 M
300 -49400 M
300 -52000 M
300 -54600 M
300 -57200 M
300 -59800 M
300 -62400 M
-7200 7200 T
R

showpage
$P e

%%Page: 12 12
/$P a D
g N
0 79200 T
S
S
7200 -74700 T
N
0 G
22233 -1050 M
/Helvetica-ISOLatin1 $
/Helvetica & P
/Helvetica-ISOLatin1 F 1000 o f
(\255 ) h
(12) h
( \255) h
300 -2290 M
-7200 74700 T
R

S
N
7500.00 -73600.00 M
53700.00 -73600.00 L
S
100 w
0 c
0 j
2 i
0.00 G k
R
R

R
S
7200 -7200 T
N
0 G
300 -1350 M
/Helvetica-BoldOblique-ISOLatin1 $
/Helvetica-BoldOblique & P
/Helvetica-BoldOblique-ISOLatin1 F 1400 o f
(2  ) h
1857 -1350 M
(Supporting Detail.) h
300 -5150 M
/Helvetica-Bold-ISOLatin1 $
/Helvetica-Bold & P
/Helvetica-Bold-ISOLatin1 F 1400 o f
(2.1 ) h
2635 -5150 M
(AI Business in Europe \(Frost and Sullivan\).) h
300 -7400 M
6000 -7400 M
/Times-Roman-ISOLatin1 $
/Times-Roman & P
/Times-Roman-ISOLatin1 F 1200 o f
31.3 0 32 (The following details are extracts from Frost and Sullivan report as presented at the) W
300 -8800 M
(Avignon Conference on AI in May 1991 by a representative of SD\255scicon \(UK\).) h
300 -11400 M
80.4 0 32 (Table 1: Comparison between market forecasts and actual sales for France, UK, Germany and) W
300 -12800 M
(Italy. \(in Millions Ecu; 1 Ecu approxs to 1.3 USD\)) h
300 -15400 M
6000 -15400 M
11700 -15400 M
17400 -15400 M
23100 -15400 M
(1988) h
28800 -15400 M
(1989) h
34500 -15400 M
(1990) h
300 -18000 M
(Frost and Sullivan forecast dated 1988 :\255) h
23100 -18000 M
(115) h
28800 -18000 M
(167) h
34500 -18000 M
(235) h
300 -20600 M
(Actual sales:\255) h
11700 -20600 M
17400 -20600 M
23100 -20600 M
(115) h
28800 -20600 M
(138) h
34500 -20600 M
(186) h
300 -23200 M
(% deviation from forecast:\255) h
17400 -23200 M
23100 -23200 M
28800 -23200 M
(17%) h
34500 -23200 M
(21%) h
40200 -23200 M
300 -25800 M
300 -28400 M
198.6 0 32 (Table 2 : Forecast for 1991 \255 1995 for the Big Four \(France, UK, Germany and Italy\) and) W
300 -29800 M
(Europe \(in Millions of Ecu\).) h
300 -32400 M
6000 -32400 M
11700 -32400 M
17400 -32400 M
23100 -32400 M
(1991) h
28800 -32400 M
(1992) h
34500 -32400 M
(1995) h
300 -35000 M
(Big Four:) h
300 -37600 M
( \255 Services) h
6000 -37600 M
11700 -37600 M
17400 -37600 M
23100 -37600 M
(137) h
28800 -37600 M
(182) h
34500 -37600 M
(426) h
300 -40200 M
( \255 Products) h
6000 -40200 M
11700 -40200 M
17400 -40200 M
23100 -40200 M
(117) h
28800 -40200 M
(162) h
34500 -40200 M
(434) h
300 -42800 M
(Total Big Four :\255) h
11700 -42800 M
17400 -42800 M
23100 -42800 M
(254) h
28800 -42800 M
(344) h
34500 -42800 M
(860) h
300 -45400 M
(Total Europe:\255) h
11700 -45400 M
17400 -45400 M
23100 -45400 M
(382) h
28800 -45400 M
(520) h
34500 -45400 M
(1286) h
300 -48000 M
300 -50600 M
(Table 3 : Euopean Vendors Market Shares.) h
300 -53200 M
(Type of Vendor) h
11700 -53200 M
17400 -53200 M
(1990 Market share) h
28800 -53200 M
(1995 Market share) h
300 -55800 M
(Computer companies ) h
11700 -55800 M
17400 -55800 M
(40%) h
23100 -55800 M
28800 -55800 M
(30%) h
300 -58400 M
(Systems integrators) h
11700 -58400 M
17400 -58400 M
(32%) h
23100 -58400 M
28800 -58400 M
(50%) h
300 -61000 M
(AI companies ) h
11700 -61000 M
17400 -61000 M
(28%) h
23100 -61000 M
28800 -61000 M
(8%) h
300 -63600 M
(Specialised niche vendors) h
17400 -63600 M
(0%) h
23100 -63600 M
28800 -63600 M
(12%) h
34500 -63600 M
-7200 7200 T
R

showpage
$P e

%%Page: 13 13
/$P a D
g N
0 79200 T
S
S
7200 -74700 T
N
0 G
22233 -1050 M
/Helvetica-ISOLatin1 $
/Helvetica & P
/Helvetica-ISOLatin1 F 1000 o f
(\255 ) h
(13) h
( \255) h
300 -2290 M
-7200 74700 T
R

S
N
7500.00 -73600.00 M
53700.00 -73600.00 L
S
100 w
0 c
0 j
2 i
0.00 G k
R
R

R
S
7200 -7200 T
N
0 G
300 -1350 M
/Helvetica-Bold-ISOLatin1 $
/Helvetica-Bold & P
/Helvetica-Bold-ISOLatin1 F 1400 o f
(2.2 ) h
2635 -1350 M
(AI Competency Circle activity FY91.) h
300 -4200 M
/Helvetica-BoldOblique-ISOLatin1 $
/Helvetica-BoldOblique & P
/Helvetica-BoldOblique-ISOLatin1 F 1200 o f
(2.2.1 ) h
3302 -4200 M
(Agenda of 29\25531 October 1990.) h
300 -6200 M
6060 -6200 M
/Times-Roman-ISOLatin1 $
/Times-Roman & P
/Times-Roman-ISOLatin1 F 1200 o f
56.4 0 32 (The first \(and last, due to the Gulf war\) meeting of FY91 focused on the following) W
300 -7600 M
(points.) h
300 -9600 M
/Symbol F 1200 o f
(\267) h
2100 -9600 M
/Times-Roman-ISOLatin1 F 1200 o f
75.1 0 32 (Discuss Customer feedback on the preceding weeks ET Seminar and agree on how/who to) W
2100 -11000 M
(follow up on leads.) h
300 -13000 M
/Symbol F 1200 o f
(\267) h
2100 -13000 M
/Times-Roman-ISOLatin1 F 1200 o f
15.7 0 32 (Discussion on AI Products Marketing strategies with Themis Papageorge, Corp. AI Market\255) W
2100 -14400 M
(ing Manager.) h
300 -16400 M
/Symbol F 1200 o f
(\267) h
2100 -16400 M
/Times-Roman-ISOLatin1 F 1200 o f
230.8 0 32 (Allow time for break\255out meetings with Corp. Marketing Manager to discuss European) W
2100 -17800 M
(needs  and perceptions.) h
300 -19800 M
/Symbol F 1200 o f
(\267) h
2100 -19800 M
/Times-Roman-ISOLatin1 F 1200 o f
(Presentation on Industrial Vision applications by Louis Pau.) h
300 -21800 M
/Symbol F 1200 o f
(\267) h
2100 -21800 M
/Times-Roman-ISOLatin1 F 1200 o f
(Presentation and discussion on AI\255IST Services by Steve Jones.) h
300 -23800 M
/Symbol F 1200 o f
(\267) h
2100 -23800 M
/Times-Roman-ISOLatin1 F 1200 o f
57.4 0 32 (GENSYM Products presentation and demo by Chuck Kiezulas, European Operations Man\255) W
2100 -25200 M
(ager  for GENSYM.) h
300 -27200 M
/Symbol F 1200 o f
(\267) h
2100 -27200 M
/Times-Roman-ISOLatin1 F 1200 o f
17.3 0 32 (Presentation and demo of practical applications of Machine Learning technology by Ed Kil\255) W
2100 -28600 M
(leen of the U.S. New Ventures group.) h
300 -30600 M
/Symbol F 1200 o f
(\267) h
2100 -30600 M
/Times-Roman-ISOLatin1 F 1200 o f
283.9 0 32 (Presentation and demo of hypermedia product, MEMEX, given by Brian Mottershead,) W
2100 -32000 M
(Product manager.) h
300 -34000 M
/Symbol F 1200 o f
(\267) h
2100 -34000 M
/Times-Roman-ISOLatin1 F 1200 o f
170.2 0 32 (Presentation of QUINTUS PROLOG \(DDS Product\) and Object Oriented Prolog by Mr.) W
2100 -35400 M
(Langoust of ELSA \(3rd party\).) h
300 -37400 M
/Symbol F 1200 o f
(\267) h
2100 -37400 M
/Times-Roman-ISOLatin1 F 1200 o f
(Presentation of ILOG Products sponsored by the French AI\255CC.) h
300 -39400 M
/Symbol F 1200 o f
(\267) h
2100 -39400 M
/Times-Roman-ISOLatin1 F 1200 o f
(Symbolic Modeling presentation given by Ty Chaney of the Corp. AITC.) h
300 -41400 M
/Symbol F 1200 o f
(\267) h
2100 -41400 M
/Times-Roman-ISOLatin1 F 1200 o f
(Presentation and demo of scheduling toolkit, given by Chris Wild of the AI\255IST.) h
300 -43400 M
/Symbol F 1200 o f
(\267) h
2100 -43400 M
/Times-Roman-ISOLatin1 F 1200 o f
79.2 0 32 (Presentation of PROLOG III \(Constraint Logic Programming,\) given by Serge Himbaut of) W
2100 -44800 M
(the AI\255IST.) h
300 -46800 M
/Symbol F 1200 o f
(\267) h
2100 -46800 M
/Times-Roman-ISOLatin1 F 1200 o f
107.2 0 32 (Discussion of how DPM could accommodate the Emerging technologies development life) W
2100 -48200 M
(cycle, with DPM Program Manager George Ristich.) h
300 -50200 M
300 -52200 M
300 -54200 M
300 -56200 M
300 -58200 M
300 -60200 M
-7200 7200 T
R

showpage
$P e

%%Page: 14 14
/$P a D
g N
0 79200 T
S
S
7200 -74700 T
N
0 G
22233 -1050 M
/Helvetica-ISOLatin1 $
/Helvetica & P
/Helvetica-ISOLatin1 F 1000 o f
(\255 ) h
(14) h
( \255) h
300 -2290 M
-7200 74700 T
R

S
N
7500.00 -73600.00 M
53700.00 -73600.00 L
S
100 w
0 c
0 j
2 i
0.00 G k
R
R

R
S
7200 -7200 T
N
0 G
300 -1200 M
/Helvetica-BoldOblique-ISOLatin1 $
/Helvetica-BoldOblique & P
/Helvetica-BoldOblique-ISOLatin1 F 1200 o f
(2.2.2 ) h
3302 -1200 M
(Report on CC Meeting of 29\25531 October 1990 \(Dirk Bade\).) h
300 -3600 M
/Helvetica-Bold-ISOLatin1 $
/Helvetica-Bold & P
/Helvetica-Bold-ISOLatin1 F 1200 o f
(2.2.2.1 ) h
4303 -3600 M
( Flags.) h
300 -5400 M
6000 -5400 M
/Times-Roman-ISOLatin1 $
/Times-Roman & P
/Times-Roman-ISOLatin1 F 1200 o f
70.2 0 32 (We have identified a serious bottleneck in corporate  AI product management: Lu\255) W
300 -6800 M
37.4 0 32 (cid Lisp V4.0 is still not available for ULTRIX V4.0. This means, we have nothing to sell with) W
300 -8200 M
8.9 0 32 (the DECstation 5xxx. VDE V2.0 is again shifted by 6 months to summer 1991. NEXPERT OB\255) W
300 -9600 M
19.4 0 32 (JECT V2.0 is ready but not delivered by Digital because N.D. didn't supply the documentation.) W
300 -11000 M
111.3 0 32 (EPITOOL will stop selling in January 1991. Only VAX OPS5 and VAX LISP \(to be retired\)) W
300 -12400 M
277.7 0 32 (are currently available. Corporate marketing now favors CSH\255agreements over new CMP\255) W
300 -13800 M
36.0 0 32 (DDS\255 or Digital products. If this trend continues we may end up with entirely different product) W
300 -15200 M
(portfolios in different countries.) h
300 -18200 M
/Helvetica-Bold-ISOLatin1 F 1200 o f
(2.2.2.2 ) h
4303 -18200 M
( HIGHLIGHTS.) h
300 -20000 M
6000 -20000 M
/Times-Roman-ISOLatin1 F 1200 o f
( The machine learning presentations & application demos ) h
34500 -20000 M
634.3 0 32 (have clearly shown a) W
300 -21400 M
85.5 0 32 (possible path to integrate AI\255components into larger applications. Machine learning itself may) W
300 -22800 M
73.1 0 32 (be seen  as a help to overcome the knowledge engineering gap for expert systems to some ex\255) W
300 -24200 M
6.7 0 32 (tent. The main message, however, is that we'll have to identify the applications that gain a com\255) W
300 -25600 M
240.2 0 32 (petitive advantage by the integration of AI\255technology and then provide more ready\255to\255use) W
300 -27000 M
147.8 0 32 (products for those possibly high volume applications. The \(unexpected\) success of our VAX) W
300 -28400 M
(Performance Advisor product may underline this statement. ) h
300 -31400 M
/Helvetica-Bold-ISOLatin1 F 1200 o f
(2.2.2.3 ) h
4303 -31400 M
( TRENDS.) h
300 -33200 M
6000 -33200 M
/Times-Roman-ISOLatin1 F 1200 o f
24.0 0 32 ( Country delegates reported significant interest in integrated configuration & sched\255) W
300 -34600 M
(uling applications. Currently the largest European AI projects with exclusive Digital staff run in ) h
300 -36000 M
(Switzerland. ) h
300 -38600 M
6000 -38600 M
19.8 0 32 (The number of dedicated AI\255specialists seems to decrease,and former AI\255specialists) W
300 -40000 M
(are now seen in other support functions \(CASE, Workstations, ULTRIX\). ) h
300 -42600 M
6000 -42600 M
31.4 0 32 (The marketing meetings aimed to find a way to extend the awareness for the highly) W
300 -44000 M
27.1 0 32 (successful AI fellowship program, which is recognized in just small but leading units in our ac\255) W
300 -45400 M
55.9 0 32 (counts, towards higher volume departmental and MIS applications. Putting AI into the account) W
300 -46800 M
59.0 0 32 (plans will fail to achieve this goal as long as we cannot offer more ready\255to\255use \(and ready\255to\255) W
300 -48200 M
(sell\) products for these applications. ) h
300 -51200 M
/Helvetica-Bold-ISOLatin1 F 1200 o f
(2.2.2.4 ) h
4303 -51200 M
( FUTURE  \(my personal view\). ) h
23340 -51200 M
300 -53000 M
6000 -53000 M
/Times-Roman-ISOLatin1 F 1200 o f
36.3 0 32 (In contrast to the former view of AI as nearly identical to expert system technology) W
300 -54400 M
23.3 0 32 (this forum has shown the move to a larger variety of possible applications. Our main concern is) W
300 -55800 M
161.9 0 32 (that AI\255projects as such are still too manpower intensive to contribute significantly to short\255) W
300 -57200 M
133.9 0 32 (term revenues. In order to push the integration of AI\255technology into higher volume applica\255) W
300 -58600 M
139.4 0 32 (tions it may become necessary to reposition ourselves to specialists for particular application) W
300 -60000 M
(domains, such as) h
300 -62600 M
/Symbol F 1200 o f
(\267) h
2100 -62600 M
/Times-Roman-ISOLatin1 F 1200 o f
( advisory systems in banking & insurance) h
-7200 7200 T
R

showpage
$P e

%%Page: 15 15
/$P a D
g N
0 79200 T
S
S
7200 -74700 T
N
0 G
22233 -1050 M
/Helvetica-ISOLatin1 $
/Helvetica & P
/Helvetica-ISOLatin1 F 1000 o f
(\255 ) h
(15) h
( \255) h
300 -2290 M
-7200 74700 T
R

S
N
7500.00 -73600.00 M
53700.00 -73600.00 L
S
100 w
0 c
0 j
2 i
0.00 G k
R
R

R
S
7200 -7200 T
N
0 G
300 -1200 M
/Symbol F 1200 o f
(\267) h
2100 -1200 M
/Times-Roman-ISOLatin1 $
/Times-Roman & P
/Times-Roman-ISOLatin1 F 1200 o f
( configuration & configuration management) h
300 -3800 M
/Symbol F 1200 o f
(\267) h
2100 -3800 M
/Times-Roman-ISOLatin1 F 1200 o f
( scheduling \(see ASF project in France\)) h
300 -6400 M
/Symbol F 1200 o f
(\267) h
2100 -6400 M
/Times-Roman-ISOLatin1 F 1200 o f
( database \(re\)design) h
11700 -6400 M
(\(see RdbExpert\)) h
300 -9000 M
/Symbol F 1200 o f
(\267) h
2100 -9000 M
/Times-Roman-ISOLatin1 F 1200 o f
(systems operation) h
11700 -9000 M
(\(see VPA, VAXsimPlus\)) h
300 -11600 M
/Symbol F 1200 o f
(\267) h
2100 -11600 M
/Times-Roman-ISOLatin1 F 1200 o f
(intelligent documentation \(see ATEM\) ) h
300 -14200 M
/Symbol F 1200 o f
(\267) h
2100 -14200 M
/Times-Roman-ISOLatin1 F 1200 o f
( intelligent document retrieval \(see EXCALIBUR\) ) h
300 -16800 M
/Symbol F 1200 o f
(\267) h
2100 -16800 M
/Times-Roman-ISOLatin1 F 1200 o f
(vision systems) h
11700 -16800 M
(\(see Louis Pau's presentation\) ) h
2100 -18200 M
300 -20800 M
300 -23400 M
300 -26000 M
300 -28600 M
300 -31200 M
300 -33800 M
300 -36400 M
300 -39000 M
300 -41600 M
300 -44200 M
300 -46800 M
300 -49400 M
300 -52000 M
300 -54600 M
300 -57200 M
300 -59800 M
-7200 7200 T
R

showpage
$P e

%%Page: 16 16
/$P a D
g N
0 79200 T
S
S
7200 -74700 T
N
0 G
22233 -1050 M
/Helvetica-ISOLatin1 $
/Helvetica & P
/Helvetica-ISOLatin1 F 1000 o f
(\255 ) h
(16) h
( \255) h
300 -2290 M
-7200 74700 T
R

S
N
7500.00 -73600.00 M
53700.00 -73600.00 L
S
100 w
0 c
0 j
2 i
0.00 G k
R
R

R
S
7200 -7200 T
N
0 G
300 -1200 M
/Helvetica-BoldOblique-ISOLatin1 $
/Helvetica-BoldOblique & P
/Helvetica-BoldOblique-ISOLatin1 F 1200 o f
(2.2.3 ) h
3302 -1200 M
(French AI Competency Circle.) h
300 -3200 M
300 -5200 M
/Times-Roman-ISOLatin1 $
/Times-Roman & P
/Times-Roman-ISOLatin1 F 1200 o f
(France has a very active CC which meets on a regular basis. Recent actions include, ) h
300 -7200 M
/Symbol F 1200 o f
(\267) h
2100 -7200 M
/Times-Roman-ISOLatin1 F 1200 o f
31.3 0 32 (A seminar combining representatives from Digital, AI Corp and ILOG held in January with) W
2100 -8600 M
(25 people attending.) h
300 -10600 M
/Symbol F 1200 o f
(\267) h
2100 -10600 M
/Times-Roman-ISOLatin1 F 1200 o f
(Initiation of press communication & writing of AI selling guide for sales & EIS people.) h
300 -12600 M
/Symbol F 1200 o f
(\267) h
2100 -12600 M
/Times-Roman-ISOLatin1 F 1200 o f
(Competency Tranfer Programs ;) h
300 -14600 M
/Symbol F 1200 o f
(\267) h
2100 -14600 M
/Times-Roman-ISOLatin1 F 1200 o f
(Avignon 91 participation.) h
300 -16600 M
6060 -16600 M
146.1 0 32 ( Problems & Support required :  \255 need more follow\255on of European AI Compe\255) W
300 -18000 M
(tency Circle meetings ) h
300 -19400 M
300 -21400 M
300 -23400 M
300 -26000 M
/Helvetica-BoldOblique-ISOLatin1 F 1200 o f
(2.2.4 ) h
3302 -26000 M
(Belgium \(Luc Vandevyver\).) h
300 -28000 M
6000 -28000 M
/Times-Roman-ISOLatin1 F 1200 o f
40.7 0 32 (There is no local AI\255CC in Belgium, the technology is covered in the CASE group,) W
300 -29400 M
17.4 0 32 (merely because the AI\255people are part of the CASE Entrepreneur Support Unit. This works fine) W
300 -30800 M
159.9 0 32 (in Belgium, since we position AI \(or ES\) as Software Engineering using other than conven\255) W
300 -32200 M
(tional techniques.) h
300 -34800 M
300 -37400 M
300 -40000 M
300 -42600 M
300 -45200 M
300 -47800 M
300 -50400 M
300 -53000 M
300 -55600 M
300 -58200 M
300 -60800 M
300 -63400 M
-7200 7200 T
R

showpage
$P e

%%Page: 17 17
/$P a D
g N
0 79200 T
S
S
7200 -74700 T
N
0 G
22233 -1050 M
/Helvetica-ISOLatin1 $
/Helvetica & P
/Helvetica-ISOLatin1 F 1000 o f
(\255 ) h
(17) h
( \255) h
300 -2290 M
-7200 74700 T
R

S
N
7500.00 -73600.00 M
53700.00 -73600.00 L
S
100 w
0 c
0 j
2 i
0.00 G k
R
R

R
S
7200 -7200 T
N
0 G
300 -1350 M
/Helvetica-Bold-ISOLatin1 $
/Helvetica-Bold & P
/Helvetica-Bold-ISOLatin1 F 1400 o f
(2.3 ) h
2635 -1350 M
(European AI Fellowship Results for FY91. \(Malcolm Reid\).) h
300 -3600 M
6000 -3600 M
/Times-Roman-ISOLatin1 $
/Times-Roman & P
/Times-Roman-ISOLatin1 F 1200 o f
221.5 0 32 (On June 28 1991, individuals from four Customer companies completed the AI) W
300 -5000 M
96.6 0 32 (Business  Fellowship Program at Sophia Antipolis, France. The four "Fellows" completed a 6) W
300 -6400 M
116.1 0 32 (week AI Training Program which started last January 15. They then worked for 4 months on) W
300 -7800 M
22.8 0 32 (Prototype Knowledge Based Systems under the guidance of  senior DEC Knowledge Engineers) W
300 -9200 M
(from the AI Group at Les Templiers \(Valbonne\).) h
300 -12400 M
/Helvetica-BoldOblique-ISOLatin1 $
/Helvetica-BoldOblique & P
/Helvetica-BoldOblique-ISOLatin1 F 1200 o f
(2.3.1 ) h
3302 -12400 M
(British Telecomm.) h
300 -14400 M
6000 -14400 M
/Times-Roman-ISOLatin1 F 1200 o f
88.5 0 32 (Mr. Mike Cox, from Network Research at British Telecom, developed a prototype) W
300 -15800 M
86.9 0 32 (which provides Network Alarm Correlation Data for Network Operations Personnel.  The pri\255) W
300 -17200 M
69.3 0 32 (mary development tool used in the Prototype is KBMS. The DEC Engineer involved was Em\255) W
300 -18600 M
23.9 0 32 (manuel Brandt. We believe the significance of the Program  lies in the fact that the BT Sponsor) W
300 -20000 M
122.9 0 32 (\( the Head of Network Research \) now  views DEC as the leader in AI applications for Net\255) W
300 -21400 M
60.7 0 32 (working. He has requested a proposal for an on\255site seminar from our group. The proposal has) W
300 -22800 M
136.9 0 32 (been  submitted. Also, BT is interested in the Fellowship "PLUS" Program which  begins in) W
300 -24200 M
(September.) h
300 -27400 M
/Helvetica-BoldOblique-ISOLatin1 F 1200 o f
(2.3.2 ) h
3302 -27400 M
(Phillips \(Holland\).) h
300 -29400 M
6000 -29400 M
/Times-Roman-ISOLatin1 F 1200 o f
39.7 0 32 (The participant from Philips Lighting, Ms. Marion Hoogenboom, developed a  pro\255) W
300 -30800 M
200.6 0 32 (totype system which generates specifications for the design of ovens which are used in the) W
300 -32200 M
5.2 0 32 (manufacture of lamps. She was assisted by Brigitte Guimbal from the Valbonne AI Group. Spe\255) W
300 -33600 M
240.0 0 32 (cific benefits for DEC as a result of the Fellowship Program are unclear at this point. As) W
300 -35000 M
16.9 0 32 (pointed out in a memo from Jan Takke, DEC Account Consultant for Philips, the  entire experi\255) W
300 -36400 M
131.5 0 32 (ence was viewed positively by Philips Management. However,  there is currently much reor\255) W
300 -37800 M
38.1 0 32 (ganization and "downsizeing" at Philips. The original sponsor for MS Hoogenboom has moved) W
300 -39200 M
127.4 0 32 (on to another position in Philips. So, the true impact of the Fellowship Program is still to be) W
300 -40600 M
(determined.) h
300 -43800 M
/Helvetica-BoldOblique-ISOLatin1 F 1200 o f
(2.3.3 ) h
3302 -43800 M
(ENEA \(Italy\).) h
300 -45800 M
6000 -45800 M
/Times-Roman-ISOLatin1 F 1200 o f
128.3 0 32 (Dr. Arcangelo Tripi, from ENEA, worked on a vision system for autonomous ro\255) W
300 -47200 M
153.1 0 32 (bots. The system will, ultimately, combine numerous advanced technologies one of which is) W
300 -48600 M
14.9 0 32 (rule based expert decision making. Dr. Trippi developed the rule based component of the vision) W
300 -50000 M
46.5 0 32 (system during the Fellowship Program. The DEC Engineer who assisted him was Carlo Taran\255) W
300 -51400 M
35.9 0 32 (tola. This Fellowship has contributed to the sales of 30 man\255years of consulting work to ENEA) W
300 -52800 M
(from DEC Italy. ) h
300 -56000 M
/Helvetica-BoldOblique-ISOLatin1 F 1200 o f
(2.3.4 ) h
3302 -56000 M
( Irish Telecomm.) h
300 -58000 M
6000 -58000 M
/Times-Roman-ISOLatin1 F 1200 o f
248.9 0 32 (The Fellow from TELECOM EIREANN was Mr. Noel Fitzpatrick. This was a) W
300 -59400 M
83.9 0 32 (unique Fellowship in that it was delivered by the Manufacturing AI Group in Galway Ireland,) W
300 -60800 M
151.9 0 32 (with Management from the AI Program Office in Valbonne. Mr. Fitzpatrick attended the AI) W
300 -62200 M
107.5 0 32 (Training Program in Valbonne, but did the prototype development at DEC Galway and at his) W
300 -63600 M
38.9 0 32 (Telecom Eireann office in Dublin. His prototype was in the area of bank cheque reconciliation.) W
-7200 7200 T
R

showpage
$P e

%%Page: 18 18
/$P a D
g N
0 79200 T
S
S
7200 -74700 T
N
0 G
22233 -1050 M
/Helvetica-ISOLatin1 $
/Helvetica & P
/Helvetica-ISOLatin1 F 1000 o f
(\255 ) h
(18) h
( \255) h
300 -2290 M
-7200 74700 T
R

S
N
7500.00 -73600.00 M
53700.00 -73600.00 L
S
100 w
0 c
0 j
2 i
0.00 G k
R
R

R
S
7200 -7200 T
N
0 G
300 -1200 M
/Times-Roman-ISOLatin1 $
/Times-Roman & P
/Times-Roman-ISOLatin1 F 1200 o f
36.4 0 32 (He worked under the guidance of Jimmy McGlennon from DEC in Galway. The immediate re\255) W
300 -2600 M
46.4 0 32 (sult is \(1\) the sale of additional consulting and project work to Telecom Eireann and \(2\) recog\255) W
300 -4000 M
154.6 0 32 (nition by credible individuals at TE that DEC should be the vendor of choice for AI related) W
300 -5400 M
(products and services. ) h
300 -8000 M
( ) h
300 -10600 M
300 -11947 M
-7200 7200 T
R

showpage
$P e

$D restore
%%Trailer
end % DEC_WRITE_dict
%%Pages: 18
%%DocumentFonts: Helvetica-Bold-ISOLatin1
%%+ Times-Roman-ISOLatin1
%%+ Helvetica-ISOLatin1
%%+ Helvetica-BoldOblique-ISOLatin1
%%+ Symbol