Information societies technology

C3.2 European Policies and Relation to Standardisation

Because of its pan-european dimension, CATEGORIES is contributing to the implementation of DG Regional Policies. Indeed the collaboration between institutions of different countries with the same objective will improve social and economic cohesion in Europe. Moreover, cooperation with European countries will not be limited to France, Spain, Netherlands or England but also with the other countries since the Network of experts in palynology will be requiring the broader and most accurate expertise. Therefore even countries engaged in joining the EU member states will br required to contribute, ensuring early integration of research teams in more institutes into the European idea.

Even if the project is ultimatelly designed to develop a generic Cognitive Vision system, it has to be stressed that the elements studied to develop and test this system will be pollen and chitinozoans. The pollen identification will allow to be alerting and preventing asthma crisis or other allergies. The chitinozoans application will have a direct impact on the oil industry as they constitute a key element in the extremely costl oil drilling strategy. A clear and automated identification of chitinozoans can therefore save substantial sums to the oil industry, making of this application a high profile support tool in determining the depth of the drilling. Therefore CATEGORIES also supports an improved quality of life and a competitive economy, two central objectives of the European Commission policies.
On top of this, the expertise gathered will be translated into computer language and processed into databases that will provide a long lasting archive for teaching, training or even run simulations or exercises. This will provide the community with the intuitive/user friendly system envisioned within CATEGORIES. The final system could indeed be of importance in palynology and geology studies as it will have integrated a wide number of varieties as well as the ways and heuristics used by experts to identify every them.

C4. Contribution to Community social objectives

The aim of CATEGORIES is to develop and advanced Cognitive Vision system. The contributioin of this system to the Community social objectives are indirect and will most likely be exprressed in the variety of applications this system will be used for. As it has been already explained, the two applications on which CATEGORIES will be relying are pollen and chitinozoans identification. Both have

C4.1 Quality of Life, Health and Safety

Life, Health and Safety

CATEGORIES will contribute to an improvement of safety of citizens through a better a monitoring of pollen rates and by defining different alert levels. The asthma disease is a significant cause of problems to an estimated 155 million people world-wide. As such it represents a major health hazard and account for substantial medical costs. In 1988, in the United Kingdom alone, these costs added up to $722 million, while indirect costs (loss of productivity, absence from work) accounted for another $81 billion. By helping prevent asthma crisis, CATEGORIES is therefore improving the living conditions of million people while avoiding extensive financial costs and losses.

Furthermore, the system can be used also for training and education. This could be performed either by consulting the database (list of pollen and ways of identification) or by running simulation on pre-selected samples. The motivation to provide better training comes directly from the ability to rely on the best updated expertise and experiences, decoded and processed into an intelligent Cognitive Vision system.

Quality of Work

Difficulties in palynology come directly from the accuracy of work and from the time required to identify a pollen grain since the moment it was collected. Some specific tolls have been developed but none of these however provides automated identification. By giving CATEGORIES the essential techniques and savoir-faire of the best specialists, the identification will not have to be done either manually or through assisting software that requires advanced training in computer science. CATEGORIES will perfom the identification automatically and get a feedback from its database from cross checking. Ultimately, the system will then enrich its database from its own analysis.

Regarding the other application, the automated chitinozoans identification will allowthe oil industry labs to perform ac much faster and reliable work. Indeed, expertise in this domain is not widely spred and the identification of the chitinozoans is relatively costly. Yet, it is unavoidable for the oil industry as the presence of chitinozoans in the undergroung layers determine the proximity of oil-fields. And when it comes to oil drilling, considering the extremely high costs, no chances are hardly taken, therefore it is essential to rely on accurate information.

In this perspective, both original applications will have valuable fall-outs in their respective domains.

C4.2 Employment

By developing a reliable Cognitive Vision system demonstrated on pollen and chitinozoans, it is expected that CATEGORIES will be modified to fit other applications. The framework produced will then help to accelerate and to enable the identification of various new elements or “objects”.

The challenge of Cognitive Vision will open up new job opportunities not only within the European research community but also within industry over the long term. Considering the huge market potential for these applications, it can be expected that the results of this pre-competitive work will be rapidly taken up by enterprises.

C4.3 Market Potential

The market potential of methodologies derived in CATEGORIES is estimated to be more than sufficient to allow the setting up specific companies. Intelligent tutorial systems, targeted application of CATEGORIES and derived applications for industry, as well as Cognitive Vision systems in industry and health-care will find numerous exploitation avenues. The development of Cognitive Vision methodologies for interpretation in new domains and the potential for these methodologies as derived in CATEGORIES is estimated to reach a minimum of 1 billion euros.