AUTOMOTIVE DESIGN
planning an AV consists of global planning – the route from point A, the beginning, to point B, the destination – alongside local path planning. In the automotive industry, engineers use both deep learning and machine learning to implement a fully automated control loop. Te control loop is comprised of three stages: motion planning, vehicular perception, and motion execution. Te loop then cycles through these stages over and over, allowing the vehicle to respond to its everchanging surroundings, and ultimately arrive at its destination safely. A control loop is defined by control logic; engineers break down the overall vehicle architecture to ensure each component is safe and meets specific safety and security standards. Simulation is then used to test and verify the safety of the control loop components by arranging various objects and events to create difficult scenarios that an AV may experience in the real world – not unlike a human driving test. Te simulation software pulls these scenarios from extensive content libraries to create absolutely any possible situation an AV
may encounter, no matter how unlikely. For example, graffiti on road signs can affect the system’s ability to read, therefore disrupting its perception.
SIMULATION AS A CATALYST TO SUCCESS If AVs are the future of the automotive industry, then simulation is the driving force behind this revolution. Within the industry, there is widespread growing pressure to make vehicle design more efficient, powerful, and reliable, all whilst reducing overall weight and size. It’s imperative that sophisticated and complex engineering trade-offs are analysed in the early stages of design, before financially investing in physical development. Based on the fundamental principles of modelling, physics, mathematics and computer science, simulation testing gives engineers the power to see how motion control capabilities will behave, and how they can be improved. Te tests provide invaluable information whilst streamlining the process of training the system’s algorithm, on the basis that these scenarios must be accurate and “true
to physics”. Te endless scenarios that simulation software can draw from ensures that vehicle testing is more thorough than ever before, all whilst being safer and more cost-effective than physical testing.
LOOKING TO THE FUTURE No longer a work of science fiction, or a figment of our imaginations, AVs now have the potential to completely transform our daily routines. As an instrumental part of the motion control process, simulation will drastically reduce the need for costly physical prototypes, allowing engineers to invest more money and resources into even greater automotive developments in the future. Simulation testing during the motion control process will not only teach and encourage AVs to reach their full potential – it will help ensure that they are safe, with the potential to save countless lives.
Dr. Christophe Bianchi is Chief Technologist – Office of the CTO at Ansys.
www.ansys.com
AV driving systems rely on sensors and embedded software for motion planning
8
www.engineerlive.com
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
