|Topic:||Implementation of ADAS feature fusion in OpenCL and performance comparison on different platforms|
|Supervisor:||Jan Šváb Garant: Doc. Ing. Tomáš Svoboda Ph.D.|
Nowadays the highend carmakers are racing towards more and more autonomous car. This trend requires more and more computing power to handle complex computations dealing with lots of sensor data. The multicore CPUs and GPUs are of course present in cars nowadays,
but up to now they have only been used for infotainment and other notsafetycritical tasks. In the very near future the market will welcome the first highend car with very powerful safetycritical ECU (Electronic Control Unit, look up Audi zFAS) which will be able to handle advanced steering maneuvers when needed to improve passenger safety and comfort. The SW development has
to address these challenges and thus the purpose of this assignment is to explore the possibilities of rather new programming approach suitable for computationally demanding tasks.
Valeo is one of the biggest tier1 automotive components supplier worldwide, active in various areas such as lighting systems, powertrain systems, comfort systems, etc. In TCE Prague our work is focussed on development of ADAS (Advanced Driver Assistance Systems) and driving automation systems.