Staff Engineer, Physical AI & Robotics Software

Renesaselectronics; Farmington Hills (MI); Hybrid; USD 150,000 - 190,000; Full time

Job summary

Renesas Electronics in Farmington Hills is seeking a Staff Engineer, Software – Physical AI to lead the adoption of Renesas R-Car SoCs for next‑gen robotic and Physical AI platforms. You will guide embedded software and AI workloads, collaborate with partners, and shape the future of R-Car products across reference designs and customer engagements.

The role requires deep expertise in embedded Linux, heterogeneous compute, and robotics software, with responsibilities spanning architecture through

Qualifications

• BS/MS/PhD in Computer Engineering, Electrical Engineering, Computer Science, Robotics, Applied Mathematics, or related field.
• 5+ years of experience in embedded systems, semiconductor application engineering, robotics, AI, autonomous systems, or high-performance compute platforms.
• Strong C/C++ software development and debugging experience.
• Deep understanding of SoC architecture, memory subsystems, firmware, operating systems, device drivers, middleware, and hardware accelerators.
• Extensive experience with Embedded Linux, Ubuntu, BSPs, bootloaders, Linux kernel, and system-level debugging.
• Strong understanding of heterogeneous compute architectures, DMA, shared memory, IPC, memory hierarchy, cache behavior, and accelerator architectures.
• Hands-on experience with GPU architectures and parallel computing.
• Experience with Vulkan, OpenCL, OpenGL ES, Vulkan Compute, OpenVX, or equivalent technologies.
• Experience profiling and optimizing system performance using tracing, logging, and debugging tools.
• Linear Algebra, Geometry & Robotics Mathematics, Optimization & Estimation, AI Computing Fundamentals.
• Ability to analyze computational complexity, memory bandwidth requirements, accelerator utilization, and data movement characteristics.

Responsibilities

• Serve as technical lead for Physical AI, humanoid robotics, autonomous systems, and robotic compute platforms based on Renesas R-Car SoCs.
• Collaborate with robotics and ecosystem partners to develop reference designs, proofs of concept, demonstrations, and reusable customer solutions.
• Support customer programs from architecture definition through platform bring-up, software integration, debugging, optimization, validation, and production.
• Analyze and optimize AI and robotics workloads including perception, sensor fusion, localization, mapping (SLAM), planning, navigation, manipulation, control, and AI inference.
• Lead system-level debugging across Linux, ROS2, AI runtimes, GPU/NPU accelerators, device drivers, firmware, memory management, IPC, and heterogeneous multicore systems.
• Optimize workload partitioning across CPU, GPU, NPU, DSP, and dedicated accelerators while balancing latency, throughput, memory bandwidth, and power consumption.
• Support GPU compute technologies including Vulkan, Vulkan Compute, OpenCL, OpenGL ES, and OpenVX.
• Develop performance models for AI and robotic systems by analyzing compute intensity, memory movement, data locality, accelerator utilization, and end-to-end latency.
• Support customer RFIs, RFQs, architecture reviews, design reviews, technical workshops, and debug engagements.
• Work with global engineering teams to define future requirements for R-Car SoCs, AI software, tools, and robotic reference platforms.

Skills

C/C++

Embedded Linux

Robotics

AI acceleration

System debugging

Education

BS/MS/PhD in a relevant field

Tools

Renesas R-Car

GPU architectures

Vulkan/OpenCL/OpenVX

Job description

Renesas Electronics in Farmington Hills is seeking a Staff Engineer, Software – Physical AI to lead the adoption of Renesas R-Car SoCs for next‑gen robotic and Physical AI platforms. You will guide embedded software and AI workloads, collaborate with partners, and shape the future of R-Car products across reference designs and customer engagements.

The role requires deep expertise in embedded Linux, heterogeneous compute, and robotics software, with responsibilities spanning architecture through