Co-operative Systems for Road Safety “Smart Vehicles on Smart Roads” SAFESPOT Integrated Project Co-operative Systems for Road Safety “Smart Vehicles on Smart Roads” Giulio Vivo Centro Ricerche Fiat SP4 - SCOVA Ljubljana, Slovenia 22 April 2008

Ljubljana, Slovenia 22 April 2008

sustainable deployment increased safety The SAFESPOT CONCEPT COOPERATIVE SYSTEMS FOR ROAD SAFETY: “SMART VEHICLES” ON “SMART ROADS” INTELLIGENT VEHICLE ADVANTAGES INTELLIGENT ROAD ADVANTAGES INTEGRATED WITH REDUCTION OF VEHICLE SYSTEM COST AND COMPLEXITY REDUCTION OF INFRASTRUCTURE COST AND COMPLEXITY INTELLIGENT VEHICLE INTELLIGENT ROAD sustainable deployment increased safety Ljubljana, Slovenia 22 April 2008

The SAFESPOT CONCEPT: from the autonomous intelligent vehicle… Ljubljana, Slovenia 22 April 2008

The SAFESPOT CONCEPT: … to intelligent Cooperative Systems Ljubljana, Slovenia 22 April 2008

SAFESPOT Integrated Project Cooperative Systems for Road Safety Project type: Integrated Project (IP) Co-funded by the European Commission Information Society and Media in the 6th Framework Programme Consortium : 51 partners from 12 European countries: OEM ( trucks, cars, motorcycles) ROAD OPERATORS SUPPLIERS RESEARCH INSTITUTES UNIVERSITIES Promoted by: EUCAR Timeframe: Feb. 2006 – Jan. 2010 Overall Cost Budget : 38 M€ (European Commission funding 20.5M€) IP coordinator : Roberto Brignolo C.R.F. Centro Ricerche Fiat (Italy) SAFESPOT Integrated Project Cooperative Systems for Road Safety “Smart Vehicles on Smart Roads” SAFESPOT is working to design intelligent cooperative systems based on vehicle to vehicle and vehicle to infrastructure communication to produce a breakthrough for road safety. SAFESPOT will prevent road accidents developing a: “SAFETY MARGIN ASSISTANT” to detect in advance potentially dangerous situations and extend, in space and time, drivers’ awareness of the surroundings. Ljubljana, Slovenia 22 April 2008

SAFESPOT Consortium Ljubljana, Slovenia 22 April 2008

SAFESPOT SPECIFIC OBJECTIVES To use the infrastructure and the vehicles as sources (and destinations) of safety-related information and develop an open, flexible and modular architecture and communication platform. To develop the key enabling technologies: ad-hoc dynamic networking, accurate relative localisation, dynamic local traffic maps. To develop a new generation of infrastructure-based sensing techniques. To develop and test scenario-based applications to evaluate the impacts and the end-user acceptance. To define the practical implementation of such systems, especially in the initial period when not all vehicles will be equipped. To evaluate the liability aspects, regulations and standardisation issues which can affect the implementation. Ljubljana, Slovenia 22 April 2008

SAFESPOT TIME FRAME 2006 Requirements 2007 Specs&development 2008 Test sites in Europe: France, Germany, Italy, Spain, Sweden, the Netherlands Technological Proto&demo Vehicles Validation On Test site Requirements 2006 Requirements 2007 Specs&development 2008 Development&test 2009 Test&evaluation Core architecture requirement Specifications &architecture Results’ Analysis Applications Integration with CVIS architecture Ljubljana, Slovenia 22 April 2008

Actual phase  Early integration Ljubljana, Slovenia 22 April 2008

DYNAMIC LOCAL COMMUNICATION NETWORK SAFESPOT KEY TECHNOLOGICAL CHALLENGES – THE VANET VEHICLE TO INFRASTRUCTURE (V2I) INFRASTRUCTURE TO VEHICLE (I2V) DYNAMIC LOCAL COMMUNICATION NETWORK the communication enables the cooperation among intelligent vehicles and intelligent roads to produce a breakthrough for road safety VEHICLE TO VEHICLE (V2V) Ljubljana, Slovenia 22 April 2008

SAFESPOT KEY TECHNOLOGICAL CHALLENGES Reliable, fast, secure, potentially low cost protocols for local V2V and V2I communication Candidate radio technology: IEEE 802.11p Need for dedicated frequency band for secure V2V and V2I, avoiding interference with existing consumer links Aligned to C2C-C and CALM standardisation groups A reliable, very accurate, real-time relative positioning A real time updateable Local Dynamic Map Ljubljana, Slovenia 22 April 2008

ACCURATE POSITIONING A reliable, very accurate, real-time relative positioning: Satellite raw data (pseudo-ranges) enhancing the proven differential procedures (DGPS). Use of landmarks registered on digital maps. DATA FUSION ALGORITHMS DGPS Vehicle sensors’ data Landmarks Other vehicles’ positions DATA FUSION ALGORITHMS DGPS Vehicle sensors’ data Landmarks Other vehicles’ positions Ljubljana, Slovenia 22 April 2008

LOCAL DYNAMIC MAPS Vehicles in queue Aim: to represent the vehicle’s surroundings with all static and dynamic safety-relevant elements Signalling phases Ego Vehicle – speed, position, status, etc Output of cooperative sensing/processing Map from provider Slippery road surface (ice) Temporary regional info Tunnel ! Landmarks for referencing Fog bank Accident (just occurred) Current concept to be refined during the project Ljubljana, Slovenia 22 April 2008

SP4 – SCOVA – Vehicle centered applications Ljubljana, Slovenia 22 April 2008

Ljubljana, Slovenia 22 April 2008

SP4 – SCOVA – Communication constraints Ljubljana, Slovenia 22 April 2008

SP4 – SCOVA – Communication constraints Control channel usage limits (US WAVE 802.11.p) rOBU = 3 Mbit/s * 580 μs / 750 ms = 2320 bit/s rRSU = 3 Mbit/s * 750 μs / 100 ms = 22500 bit/s Applicative models where “actor a ask actor b for…” can not be adopted due to the implications related to the poor usage of the communication channel. Ljubljana, Slovenia 22 April 2008

This implies that a rigid synchronization between the world representation (LDM content) of the different vehicular nodes of the VANET can not be achieved. The single parameter enforcing the coherence between the different representation is the time, which is common and shared. So, dedicated tasks (both at the network level and at the applicative level) should be implemented for achieving a good “alignment” between the different LDMs. Ljubljana, Slovenia 22 April 2008

An example of the reference model that has been specified is related to the Speed limitation & safety distance application. Let’s say that actor 2 needs to know the status of the brake pedal of actor 1. This parameter is easily available on the CAN bus of 1, but this information should be transmitted only after becoming aware of belonging (with a passive participation) to an applicative scenario where the assistance effects - warnings - are exclusively for the benefit of 2. Ljubljana, Slovenia 22 April 2008

Ljubljana, Slovenia 22 April 2008

The two state machines running on the primary actor and on the secondary actors of any given scenario are based on the VANET beacons; this should allow to build up a similar representation of the surrounding scenario Ljubljana, Slovenia 22 April 2008

An Application Manager, running on the primary actor; At the end, four applicative tasks are running in order to support the cooperative applications: An Application Manager, running on the primary actor; A Driver Assistance Application, running on the primary actor; A Message Manager, running on each secondary actor; A Cooperative Assistance Application, running on the secondary actors. Application Manager + Driver Assistance Application Message Manager + Cooperative Assistance Application Ljubljana, Slovenia 22 April 2008

Conclusions An overview of the SAFESPOT project has been presented. Benefits of the cooperative approach are ranging from the sharing low-level data to the usage of integrated information for implementing a novel Safety Assistance for the users of the equipped vehicles. At the applicative level the advantages of using a cooperative approach are also significant; the complete situation around the vehicle can be described within a local dynamic database (the LDM) where some safety critical parameters (as the minimum positions and dynamics of the VANET actors) are shared by all vehicles in the surrounding. Vehicles can exchange their reciprocal position and negotiate for lane changes; dynamic hot spots information can be propagated among the co-operating vehicles. This level of information would be especially useful for the lesser equipped vehicles that do not carry full ADAS equipment but only have available some basic intercommunicating SAFESPOT unit. Finally, unequipped vehicles may also benefit as some information can be transferred from other SAFESPOT actors to traffic control centers and distributed by means of traditional medias such as information on the public radio or Variable Message Signs. Ljubljana, Slovenia 22 April 2008

REFERENCES IP web site www.safespot-eu.org IP Coordinator Roberto Brignolo, Centro Ricerche Fiat Tel. +39 011 9080 534 safespot@crf.it Core Group Centro Ricerche Fiat, Renault, Volvo, DaimlerChrysler, Magneti Marelli, Bosch, COFIROUTE, ANAS, TNO Sub-Projects Leaders SP1 – SAFEPROBE Christian Zott, Robert Bosch GmbH, christian.zott@de.bosch.com SP2 – INFRASENS Angela Spence, MIZAR Automazione, angela.spence@torino.miz.it SP3 – SINTECH Achim Brakemeier, DaimlerChrysler, achim.brakemeier@daimlerchrysler.com SP4 – SCOVA Giulio Vivo, Centro Ricerche Fiat, giulio.vivo@crf.it SP5 – COSSIB Guy Fremont, Cofiroute, guy.fremont@cofiroute.fr SP6 – BLADE Han Zwijnenberg, TNO, han.zwijnenberg@tno.nl SP7 – SCORE Abdelkader Mokaddem, Renault, abdelkader.mokaddem@renault.com Quality&Dissemination, Angelos Amditis, ICCS, a.amditis@iccs.gr Ljubljana, Slovenia 22 April 2008