From patrick.schouller@inria.fr Fri Mar 19 08:20 MET 1999 To: (Recipient list suppressed) From: Patrick SchoullerSubject: ALGORIEL Consulting propose a partnershaft to experiment a new management information system in enterprise MEFI-SEI-DiGITIP-STSI-SDPP-PS-1999-106 PROJECT PROPOSAL TENTATIVE TITLE OF THE PROJECT: MPI METHOD : Modelling of management information system and identifying software components for reusable software. CONTACT PERSON & CONTACT INFORMATION: Name : Georges JACOVLEV Organisation : ALGORIEL Department : ALGORIEL Consulting Mailing adress : Tour Montparnasse, 33, avenue du Maine, 75755 PARIS cedex 15 - FRANCE Telephone : (33) 1 45 38 36 05 Fax : (33) 1 45 38 36 22 URL : www.algoriel.fr Email : georges.jacovlev@wanadoo.fr PROJECT DESCRIPTION: We have already experimented our method, called MPI (Modelling of Processes through Information), in different enterprises : FRANCE TELECOM, L'AIR LIQUIDE, MERLIN GERIN, VERRERIES CRISTALLERIE D'ARC, VIDEOTHEQUE DE PARIS, LA POSTE, IBM FRANCE, BNP, R.T.L., BANQUE DE FRANCE, SNCF. But these were only partial experimentation and, what concerns reusable software was in a very small way experimented, but sufficient to know that it's operative. Now, we want to experiment the whole method in an enterprise, in each country of our partners, to demonstrate the benefits it brings, particularly : - to know really what are the acting processes that contribute to the finality of the enterprise; - to be able to build an information system that reflects the real operations, and not what the I.T. technicians want to build, so that the information system could bring a real aid to the management of the enterprise; - to be able to build industrialy, with a low cost, reliable and flexible software applications. We know that we are able to split a complex organisation or system into, for example, 150 or 200 separate functional objects. Not a method in the world can do it when the system is complex. We know that we are able to build a big software application by fitting together (probably) no more than 200 software components. The different experimentations of reusable software we met have built a minimum of five thousands of software objects. We think it's too much to be easy reused. The fundamentals of our method MPI are : - identify a "conceptual langage", different of the business language and of the I.T. language, - be able to solve the contradictions that are found during the usual ways of organisation modelling, - measure the information quantities that are in the processes, that determines two parameters - "utility" and "complexity" - for each process and for each step of each process, - identify the elementary functional objects that exist in each step of each process, that is the way of building reusable software. How we propose to proceed : - first we shall teach our partners how to use our MPI method, at the beginning of the research, and during it, as necessary, - then the partner, if it is a software engineering or consultant firm, has to find a pilot enterprise in which the experimentation could be done; another way is that an industrial or service enterprise is directly interested in the experimentation. EXISTING PARTNERS: We don't have any final agreements with our potential partners, yet. Negotiations has started, abroad with software engineering and consultant firms and, in France with industrial and service enterprises. **************************************************************** Patrick SCHOULLER MINISTERE DE L'ECONOMIE, DES FINANCES ET DE L'INDUSTRIE SECRETARIAT D'ETAT A L'INDUSTRIE Direction Generale de l'Industrie, des Technologies de l'Information et des Postes Service des Technologies et de la Société de l'Information 3/5 rue Barbet de Jouy 75353 PARIS 07 SP Telephone : + 33 1 43 19 34 25 Telecopie : + 33 1 43 19 35 51 Archive de la liste : http://www.evariste.org/actu/schouller/ http://www.industrie.gouv.fr. http://www.telecom.gouv.fr ****************************************************************