~/research/samlei
Samuel Leitenmaier
Research Associate · TH Augsburg
I work at the intersection of machine learning based planning and hardware acceleration for autonomous vehicles — with a particular interest in heterogeneous SoCs can make real-time autonomy both efficient and reliable.
active research
pursuing PhD
looking ahead
// PhD candidate (currently)
Working toward a doctorate on planning-oriented autonomous driving beyond observed futures.
research interests
- Autonomous driving — motion planning & scene understanding
- World models
- Heterogeneous computing with FPGAs and SoC platforms
- Hardware/software co-design for real-time robotics
- Edge inference on resource-constrained embedded systems
publications
Towards Specialized Hardware for Autonomous Driving Functions: THAccelerated Motion Planner
↗ abstract
Autonomous Electric Race Car Inverter Development: Revving up the future with resource efficient drive technology
↗ ieee xplore
projects
NeMo.bil
Reponsible for technical coordination and system development of the autonomous driving system. Second university to get L4 testing approval on public streets.
THAMP
FPGA-accelerated motion planner for autonomous vehicles, validated in hardware-in-the-loop and real-world testing.
YOLOv4 Edge
Object detection comparison between Jetson TX2 and UltraScale MPSoC — notable power efficiency advantage on FPGA.
FS Inverter
Traction inverter for Formula Student EV — EnDat encoder interface, CAN-AXI bridge, custom embedded Linux stack.
tools & languages
VHDL
HLS
C/C++
Python
Xilinx Vivado
ROS 2
Embedded Linux
Zynq UltraScale
CAN Bus
U-Boot
contact
- email[email protected]
- websamlei.io
- cvsamlei.io/cv