AUTOMOTIVE TESTING
Swarm testing cuts development time
complex scenario testing can be transferred from one environment to the other with centimetre accuracy. This enables the testing and development of autonomous and advanced driver assist systems (ADAS) to be significantly accelerated whilst also reducing risk and cost. As the industry continues to move towards autonomous technology, the complexity of testing is also increasing. AB Dynamics coordinates a group of vehicles (a swarm) in a test that uses the company’s vehicle control robots and wireless telemetry system. In this way, the vehicles are un-manned and communicate both with each other and with the control centre. “We are the only company to use the same
A
software tool chain for simulated and real-world testing; this means we can develop and run a test in the virtual world and simply ‘copy and paste’ it when at the test track,” says Jeremy Ash, sales director at AB Dynamics. “Swarm testing is very complex and can involve the precise choreography of eight or more test items. Our system enables customers to run thousands of iterations of tests virtually and then confidently replicate a sample of these in the real world for correlation.”
Scenario testing using a swarm of vehicles is a vital aspect of development for ADAS and autonomous vehicle technologies. It assesses the ability of autonomous functions to interpret the behaviour and intentions of drivers and vehicles in the close vicinity. Scenarios such as merge-in, caused by lane closures on dual-carriageways, cut-in and cut-out are common events that force a vehicle using ADAS or autonomous systems to make a decision (brake, merge, accelerate or perform an evasive manoeuvre).
8 /// Testing & Test Houses /// March 2021
❱❱ Swarm testing aims to replicate complex interactions between vehicles in close proximity, above; test vehicles are equipped with sophisticated control systems, inset
utomotive testing specialist, AB Dynamics has been demonstrating how simulated swarm tests can be replicated quickly and accurately in the real world and vice- versa. By using a common tool chain,
Complex testing scenarios can be taken from the simulator to the track using the same software
“Thousands of tests are required to establish an autonomous vehicle’s ability to safely navigate such scenarios,” continues Ash. “The use of a single tool chain enables the bulk of these tests to be conducted in simulation – reducing test track costs and the reliance on engineering time. Using the same software that controlled the simulated world, test parameters are essentially lifted to conduct the same tests on the track. This correlation provides the necessary confidence to accelerate simulated testing and bring next-generation technology and functionality to market more efficiently.”
PROOF OF CONCEPT AB Dynamics recently completed a demonstration with VW Group using seven automated vehicles, which were synchronised with an Ego vehicle (the vehicle under test) in a range of trials at the Nardò Technical Centre in Italy.
A total of 152 tests were carried out over 11 days, including high-speed runs up to 130kph and “cutting in” manoeuvres with vehicles as close as 1m apart. AB Dynamics supplied test equipment and engineers from its UK and German offices, and from its Italian distributor, Leane, which worked in conjunction with personnel from VW, Audi and Porsche. “These tests were much more complex than a Euro NCAP scenario, which typically involves one target vehicle plus a guided soft target,” explains Leo Evans, senior engineer at AB Dynamics. In the real world, interactions often involve multiple vehicles but running several automated vehicles in close proximity has challenged the industry until now, meaning human drivers were often used instead. “Unfortunately, human drivers, however highly skilled, cannot always produce the repeatability needed to thoroughly investigate system behaviour, so electronic control is necessary to eliminate variation between tests,” concludes Evans. T&TH
Page 1 |
Page 2 |
Page 3 |
Page 4 |
Page 5 |
Page 6 |
Page 7 |
Page 8 |
Page 9 |
Page 10 |
Page 11 |
Page 12 |
Page 13 |
Page 14 |
Page 15 |
Page 16 |
Page 17 |
Page 18 |
Page 19 |
Page 20 |
Page 21 |
Page 22 |
Page 23 |
Page 24 |
Page 25 |
Page 26 |
Page 27 |
Page 28 |
Page 29 |
Page 30 |
Page 31 |
Page 32 |
Page 33 |
Page 34 |
Page 35 |
Page 36 |
Page 37 |
Page 38 |
Page 39 |
Page 40 |
Page 41 |
Page 42 |
Page 43 |
Page 44 |
Page 45 |
Page 46 |
Page 47 |
Page 48 |
Page 49 |
Page 50 |
Page 51 |
Page 52
