Efficient Implementation of Parallel Path Planning Algorithms on GPUs

Abstract

In robot systems several computationally intensive
tasks can be found, with path planning being one of them.
Especially in dynamically changing environments, it is difficult to
meet real-time constraints with a serial processing approach. For
those systems using standard computers, a promising option is to
employ a GPGPU as a coprocessor in order to offload those tasks
which can be efficiently parallelized. We implemented selected
parallel path planning algorithms on NVIDIA's CUDA platform
and were able to accelerate all of these algorithms efficiently
compared to a multi-core implementation. We present the results
and more detailed information about the implementation of these
algorithms.