RoadNex detects free space on road with negative detection of obstacles as vehicle on the video below.
Nexyad was present with Groupement ADAS at Equip’Auto Congress in Paris. Groupement ADAS is a SME’s cluster : 10 companies with expertise in the field of Advanced Driver Assistance Systems, Connected car and Autonomous vehicle. Philippe Orvain CEO of Nomadic Solutions and competitiveness cluster MOV’EO Vice President has responsed to journalist Laurent Meillaud on Congress TV channel.
Watch Philippe Orvain interview on the congress channel with SafetyNex video demo :
Nexyad CEO Gérard Yahiaoui A.I. expert was invited to speak (invited paper) in a conference in Paris on Artificial Intelligence (JNI/IESF, Under the patronage of UNESCO, 2017 Oc 19). He talked about Artificial Intelligence for Autonomous Vehicle : “Intelligence Artificielle pour le Véhicule Autonome, et exemple de réalisation : SafetyNex”. Very interesting papers during this day showing a broad range of AI applications (automotive, fashion, HR management, Legal, …). More than 400 people in the audience.
Come to visit Groupement ADAS booth with NEXYAD Automotive & Transportation at Equip’Auto Symposium in Paris, Porte de Versailles.
The NEXYAD company is currently developing the construction of a database for the validation of systems of driver assistance and driving delegation, (ADAS and Autonomous car) using the AGENDA methodology published in the 1990s by Gérard Yahiaoui (methodology initially intended to handle, among other things, the construction of learning databases and tests for the implementation of neural networks).
This database has two essential characteristics:
1) Real-life situations
Indeed, the AGENDA methodology recommends describing the possible variations of signals and input images as factors of variability and their crosses.
Example, for obstacle detection:
. weather (dry weather, sunny weather, rain, fog) . overall brightness (low, medium, high)
. vehicle speed (low, moderate, high)
. type of road (motorway, road with marking, road without marking, …)
. coating (bitumen 1, bitumen 2, …, pavers)
. day / night (car headlights and infrastructure lighting)
. season (spring, summer, autumn, winter)
Type of obstacle:
– static obstacle
. linked to the infrastructure: works terminals, tolls, …
. related to users: tire on the roadway, package dropped from a truck, motorcyclist lying on the road following an accident, vehicle broken down stopped on the roadway, pedestrian stationary on the edge of the roadway (visible / hidden)
– moving obstacle
. truck, car, vulnerable (pedestrian, bike, motorcycle) with the typical trajectories (longitudinal in the direction of travel, longitudinal in the opposite direction of travel, lateral) and the position (opposite, right, left) .
Real-life situations examples :
We see that if we combine these factors, we can find quite quickly a very large number of cases. Now, the development of ADAS systems is complex, and it is necessary to proceed by successive iterations, from simple situations to complicated situations.
Our database allows this, since all records are described such as cross-referring of modalities of the factors of variability. We thus know exactly in which cases the system was tested or not.
The formalism of ‘cross-referring of modalities of the factors of variability’ makes it possible to use experimental designs, and in particular orthogonal fractional plans, to greatly reduce the number of cases to be tested while guaranteeing maximum coverage of real-life situations. In this context, we can develop an ADAS on an orthogonal fractional plane and test it for other orthogonal fractional planes, for example.
2) Ground reality
It is a question of define the obstacles and elements of the infrastructure (markings, road edges, etc.) on the images so as to constitute a reference allow measuring the performance of the system.
1.1, summer, cloudy, unmarked road, moderate speed, tire on roadway, dry weather
1.2, summer, cloudy, unmarked road, moderate speed, package on roadway, dry weather
2.1, summer, cloudy, unmarked road, moderate speed, visible static pedestrians at the edges of the roadway, time
2.2, summer, cloudy, unmarked road, moderate speed, human lying on roadway, dry weather
It is not certain that one can meet these few cases, even by driving 1 million km on open road!
Target of this database :
NEXYAD starts its collection of images and data:
. video (towards the front of the vehicle) color
The files are synchronized by the RT-MAPS tool of the company INTEMPORA.
The files are saved in RT-MAPS format and directly replayable on this tool.
NEXYAD is currently looking for contributors on this internal project. Contributors co fund and have in return free access to the database, unlimited in time. This contribution will accelerate the work of collection and labeling.
NEXYAD wishes to make this database available soon, free of charge to give material to the ADAS community and the autonomous vehicle, in a reduced version of the database, and in a paid way (in the form of subscriptions) for the complete base.
NEXYAD’s ambition is to propagate its methodological expertise and to enable everyone to evaluate the performance of vision systems for ADAS, be they systems developed by NEXYAD or others.
“Methodology for ADAS Validation: The Potential Contribution of Other Scientific Fields Which Have Already Answered the Same Questions”, Gérard Yahiaoui, Pierre Da Silva Dias, CESA congress Dec 2014, Paris, Proc. Springer Verlag
As many people know now, NEXYAD has been developing the first real time driving risk assessment system called SafetyNex.
SafetyNex is currently available in B2B :
. as a smartphone App (Android and IOS)
. as a real time driving risk assessment API that OEMs and Insurers may integrate into their own smartphone App or into their own telematics or ADAS device (Android, iOS, Linux, Windows).
This real time driving risk assessment module has been validated on 50 million km, and applies proven methods for risk assessment, using, for instance, the Frank E. BIRD “safety triangle” concept, and running in real time a knowledge-based system AI that has been built by NEXYAD since 2001. It took 15 years to extract thousands of road safety knowledge atoms from experts of 19 countries. Some of this knowledge is directly operational, some is deep knowledge on detection theory (a mix of Information Theory and Knowledge on Human Brain abilities). And of course, SafetyNex also applies fundamental knowledge on mechanics (braking abilities, …) including complex issues such as grip for example.
A tough work was to be made in order to build such a knowledge-based system that could cope with “gradual truth” and could run a complex knowledge-based model in real time. Implementation uses fuzzy sets both for facts and knowledge representation, and competition between knowledge atoms uses deep Possibility Theory concepts. Of course, SafetyNex deals with dynamic problems (road accident, with distances, speed, …) and had to find a proper representation of TIME. NEXYAD was not very quick to develop all those solutions and to integrate them into a product called SafetyNex : SafetyNex is proposed for sale since Jan 2017 (first R&D on SafetyNex : 2001). But the result is efficient : you can run SafetyNex on your smartphone while driving your car and see how it helps you to drive much safer without alerting for nothing.
This unique technology places now NEXYAD in the center of the worldwide telematics/adas/data collecting race for car insurers.
So let’s see what are the main applications of SafetyNex for car insurers :
. PREVENTION : SafetyNex alerts (vocal alert) driver BEFORE a dangerous situation, letting time to slow down and avoid accident. It is the best prevention tool ever : studies showed that 20% of accident can be avoided (then also 20% of personal injuries and fatalities !!!). SafetyNex is the ONLY tool that runs in real time inside the smartphone or inside the telematics/adas devices and then is the ONLY module that can alert driver (all the other systems apply scoring methods on the cloud …). This is a major differenciation for SafetyNex : avoid accident ! Observation of drivers using SafetyNex showed that they change their way of driving in order to mimnimize the number of alerts, then they change their driving behaviour in the long term.
. COMMUNICATION : SafetyNex gives driver a score of Safe Driving after every trip. This is a positive way of presenting risk and driver appreciate being given a gold cup for his/her high safe driving score. You can imagine many ways to communicate for car insurers. Example : every year, sort all drivers (customers) of a country, find the best (highest SafetyNex safe driving score) and give a huge incentive (offer a brand new car, for instance). Such an incentive tranforms the car insurer’s customers set into a community that tries to drive better, and you can build of course a social network for experience share and local challenges … etc. This is a simple and very concrete use of digital technology both for getting better profits and for saving lives. Marketing executives call this ‘the Nudge Theory”. SafetyNex is a natural candidate for applying this Nudge Theory.
. KNOWLEDGE : car insurers only know about risk by accident observation. Although the theory of Frank E. BIRD (safety triangle) shows that the proper variable to observe would be what is called “near miss” or “quasi accident”. SafetyNex alerts every time that driver’s behaviour leads to a quasi accident situation. Then car insurers can use SafetyNex to build this new knowledge and sort and cross all their variables with this new notion. Actuaries will quickly find clever ways to use those new sets of data and you will be surprised to see how productive data scientists can be when they have a direct access to new qualified huge set of data.
. USAGE BASED INSURANCE : Of course, SafetyNex is THE perfect tool for UBI. At the end of every trip, SafetyNex sends data (statistics) : usage (urban, road, highway, time, duration, km, postal code, etc …) and risk (risk profiles). Then it is very easy to propose Usage Based princings (‘as you drive” or “how you drive” etc).
It is important to notice that real time abilities of SafetyNex also simplify business models :
. Anyway car insurer does effective prevention with SafetyNex : it is only needed to avoid few personal injuries and fatalities to “pay” the deployment of SafetyNex. That’s the key of SafetyNex deployment : SafetyNex alerts driver WHILE DRIVING and BEFORE danger, then many accident are avoided … then many personal injuries are avoided. If you know the price of a personal injury in you car insurance, then you see that SafetyNex price was made in order to let you get a fair ROI on this issue.
. Communication (incentive) is then already paid by the number of avoided personal injuries !
. and in addition, car insurer gets qualified data : for knowledge acquisition and UBI opportunities study.
As a conclusion, we can remember main facts :
. SafetyNex real time abilities make it VERY DIFFERENT from any other risk assessment solutions (using stats and data science in the cloud) because then SafetyNex ACTS on driving behaviour BEFORE accident and then lets avoid many accidents.
. Prevention effect of SafetyNex finances it deployment cost.
. Knowledge extraction and UBI opportunities business models can be done without stress as SafetyNex already gives a fair ROI to car indurer.
Some OEMs and insurers already started to integrate SafetyNex API into their smartphone App or into their telematics/adas device.
It is now time to move forward to the 21st century with digital insurance.