The 3xD Simulator for Intelligent Vehicles Professor Paul Jennings 20 th October 2016
An academic department within the science faculty Established in 1980 by Professor Lord Bhattacharyya as Warwick Manufacturing Group to facilitate technology transfer and knowledge creation for Industry 500+ people (800+ university and industry) working in 7 buildings Training over 1,500 individuals in the UK and abroad (from school to post experience) Professor Lord Bhattacharyya Founder and Chairman of WMG Co-located with Jaguar Land Rover & Tata Motors European Technical Centre
NAIC: National Automotive Innovation Centre 150m investment, led by private sector National focus of automotive research capability An environment to foster collaboration, cohesion and cross-fertilisation of knowledge Academic, industry (including supply chain) teams 33,000m 2 building: Design studio & support spaces Virtual engineering areas Powertrain/Hybrids dyno hall Multi-purpose drive-in engineering bays Electronics and other labs Hands on teaching areas Office space for 1,000 Collaboration space Substantial exhibition area Café Greener, lighter, smarter vehicles
Intelligent vehicle technology will bring benefits for all of us Why? 90% of current accidents are due to human error Improved energy efficiency and air quality Reduced congestion Greater productivity e.g. average UK driver spends 235 hours behind the wheel Independent mobility for all Huge new business opportunities for many sectors But we need to ensure that vehicles will be safe, secure and robust
Real World Trials UK is encouraging national use, so our experience informs international scene ( ISO, type approval etc.. ) Government has published a Code of Practice Important for public engagement
Trials are hugely valuable but We need to minimise risk There is no repeatability, and limited control of the test conditions High cost & low efficiency
How many miles does it take to test an autonomous vehicle? Karla N & Paddock S 2016 Driving to Safety: How Many Miles of Driving Would it Take to Demonstrate Autonomous Vehicle Reliability RAND Corporation
Trials are hugely valuable but We need to minimise risk There is no repeatability, and limited control of the test conditions High cost & low efficiency We also need to bridge the gap between real vehicle demonstration and computer simulation, whilst still involving real people. New methods are required (amongst a continuum of approaches)
WMG 3xD Simulator for Intelligent Vehicles Vision: To test or evaluate any new technology (infrastructure, communications and on-vehicle) in representative real world conditions with a driver in the loop
WMG 3xD Simulator for Intelligent Vehicles 360 degree cylindrical screen, HD projection, surround sound and eye tracking Vehicle motion Vehicle control systems emulation to trick the car into thinking it is running Emulation of external wireless signals such as GPS, ITS, 3G/4G and 802.11p 30 mile LiDAR scanned and graphically rendered driving route urban, intra-urban and motorway routes around Coventry
Our early research focus Testing and evaluation Virtual design verification process and real time testing Functional safety & ISO26262 Use Case and Scenario Definition Driver in the Loop Cybersecurity Perception, acceptance and usability (user trials, HMI, sound quality, driver monitoring, distraction, user-centric design) Communications technology New business opportunities Current projects include: INTACT, UKCITE, VVIL and 8 doctoral projects
WMG s CAV1 projects INTACT: INnovative Testing of Autonomous Control Techniques
INTACT INnovative Testing of Autonomous Control Techniques Virtual Lidar sensors Needed for the base map Recreate dynamic LiDAR beam data Virtual camera Needed for object detection Recreate separate video images Options Computer vision approach Use of server-client data from simulator Extract video from simulator in real-time
Advanced scenario and test case generation VVIL Constrained randomised testing Case study: V2V co-operative driving Starting scenario number of vehicles speed and path of vehicles braking/acceleration of vehicles environmental conditions Runtime interaction with the simulator apply driving input receive data as if from sensors Tackling the Autonomous and Connected Vehicle Testing Challenge - Ross Addinall, Vertizan Vertizan Limited
Testing continuum Digital World - Simulation 3xD Simulator Track Testing Real World Trials Realism Control of operating environment User-in-the-loop Increasing ability to choose scenarios Flexibility to test different technology Dependence upon real system availability Faster than real-time Efficiency Real-time Aim: to confidently test earlier in product development
Summary New technology and vehicles are coming sooner than we might think, bringing benefits for all of us. But we need to ensure they are secure, safe and robust in complex real world environments To do this, we will need new infrastructure, real world trials, and new methods too. WMG is developing a new and unique capability for virtual prototyping, to reduce R&D costs and accelerate commercialisation
Contact Details Paul.Jennings@warwick.ac.uk (WMG Lead for Connected & Autonomous Vehicles)