Artificial intelligence: 15 New Projects Will Be Tested In Administrations
Facilitate medical decisions, fight against false customs declarations, improve coastal mapping… 15 new concrete applications of artificial intelligence will be tested within the administration. These projects were chosen from among the 46 proposed by administrations during the second call for expressions of interest Artificial Intelligence , launched last April by the interdepartmental directorate for digital and the State information and communication system (DINSIC) and the interdepartmental direction of public transformation (DITP).
 |
| Applications OF Artificial Intelligence |
Prototype and business support
Financed up to 4M € as part of the investment program for the future, these 15 projects will be supported from September 2019, for one year, until the production of a functional prototype, tested with users. DINSIC will provide technological expertise in data processing, artificial intelligence and more broadly digital transformation. In parallel, the DITP will support each administration to identify upstream, prepare and then support the business transformations induced by the integration of AI in the processes.
AI, a future challenge for public service
These 15 new projects are in addition to the 6 prototypes under development, resulting from the first call for expressions of interest . Together, they allow the administration to test the potential of AI in the field, through concrete applications . By freeing agents from certain tasks, these technologies will allow them to refocus on their core business, in particular through more personalized support for users, for better service.
The 15 winning projects
Optimizing the processing of reports of adverse health events
General Directorate of Health
The data used
All reports posted on the portal signalement-sante.gouv.fr
Artificial intelligence
Artificial intelligence can automatically classify reports, extract relevant business concepts and detect anomalies in reports.
Gains
- Evaluators' work made easier
- Improved return times to the declarant
Facilitate access to patient information
Bordeaux University Hospital Center
The data used
The computerized patient record
Artificial intelligence
Artificial intelligence will allow a semantic search in the computerized patient record to find the right information at the right time.
Gains
- Work of nursing staff made easier
- Save time searching for patient information
Develop an automatic alert system to help monitor workers exposed to ionizing radiation
Institute for Radiation Protection and Nuclear Safety
The data used
The database of the information system for monitoring exposure to ionizing radiation (SISERI)
Artificial intelligence
Artificial intelligence algorithms will allow personalized monitoring of exposure to ionizing radiation and easier detection of anomalies.
Gains
Better protection of workers exposed to ionizing radiation
Develop artificial intelligence in dematerialized legality control
Directorate General of Local Authorities
The data used
Acts transmitted by local authorities to prefectures
Artificial intelligence
By automatically sorting the transmissible and non-transmissible acts and detecting the information to be checked as a priority, artificial intelligence can facilitate the work of agents in the prefecture.
Gains
Work of agents in simplified prefecture
Improve the online pre-complaint system
National Gendarmerie Directorate
The data used
Online pre-complaints
Artificial intelligence
By automatically detecting infringements from online pre-complaints and identifying additional questions to ask the user, artificial intelligence will make it possible to transform online pre-complaints into qualified complaints.
Gains
Save time for both complainants and agents
Identify false customs declarations using text mining
Directorate General of Customs and Indirect Rights
The data used
Unique administrative documents
Artificial intelligence
From the descriptions of the products in the single administrative documents, artificial intelligence will make it possible to detect imported products which are wrongly declared in a nomenclature of products that is fiscally advantageous.
Gains
Reduction in customs fraud
Improve the identification of the employer establishment in the population census
National Institute of Statistics and Economic Studies (INSEE)
The data used
The declarations of employing establishments in the population census and the SIRENE database of establishments
Artificial intelligence
Artificial intelligence will help develop a recommendation engine for officers correcting census data.
Gains
- Greater precision of home-work statistics provided by INSEE
- Increased efficiency in the data production chain
Develop a tool to help the establishment of craft businesses
Chambers of trades and crafts
The data used
Data on artisans and cross-referenced data
Artificial intelligence
Artificial intelligence will determine the chances of success of a craftsman in a territory.
Gains
Choice of location of more enlightened artisans
Identify molecules contaminating the environment and profile sources of pollution
National Institute for Industrial Environment and Risk
The data used
Results of chemical analyzes
Artificial intelligence
Artificial intelligence will make it possible to objectify similarities and differences between molecular spectra to better identify the molecules contaminating the environment.
Gains
Better detection of polluting molecules
Develop a monitoring and information exploitation solution to improve economic security
Directorate-General for Enterprise - Strategic Information and Economic Security Service
The data used
Foreign investment requests in France crossed with external sources
Artificial intelligence
Artificial intelligence will make it possible to detect and identify entities presenting a risk to the economic interests of France.
Gains
- Improve the processing of economic intelligence files
- Better detect the risks that foreign investments can present
Automatically identify the series of disputes calling for a joint decision
Board of state
The data used
All applications to administrative courts
Artificial intelligence
Artificial intelligence will allow litigation to be grouped to automatically detect series.
Gains
Better efficiency of justice
The following 4 more complex projects will benefit from scientific support in collaboration with a research institute:
Detect unfair terms and practices in contracts, quotes and invoices
Directorate General for Competition, Consumer Affairs and Fraud Prevention
The data used
The reports and documents recovered by the DGCCRF investigators during the controls as well as the decisions, opinions and recommendations of the unfair terms committee
Artificial intelligence
Artificial intelligence will make it possible to detect the presence and content of mandatory information, to identify the themes of contract clauses and, if necessary, to identify unfair clauses.
Gains
- Better consumer protection
- Better competitive market regulation
Facilitate the calculation of the positioning of terrestrial reference points from satellite data
National Institute of Geographic and Forest Information (IGN)
The data used
The positions of the permanent antennas
Artificial intelligence
Artificial intelligence will automatically identify breaks in the measurement of the positions of the permanent antennas in order to be able to correct them.
Gains
Improving the quality of the international terrestrial reference system
Identify differences in case law
Court of Cessation
The data used
Jurisprudence from the Court of Cessation (Jurinet) and the courts of appeal (JuriCA)
Artificial intelligence
Artificial intelligence will bring together decisions dealing with the same subject and detect divergent interpretations of the law.
Gains
Better guarantee of a uniform interpretation of the law
Improve coastal mapping
Hydro-graphic and Oceanographic Service of the Navy
The data used
Laser remote sensing (lidar) measurements taken on the coast
Artificial intelligence
Artificial intelligence will improve data production by learning from corrections made by agents to automate part of the process.
Gains
- Improvement of the quality and accuracy of the coastal reference system
- Agent work made easier
No comments: