Lead Edge Architect
-
Software Development
Job Description & How to Apply Below
Lead Edge Architect in the perception and edge function, covering the on-machine compute platform for data collection and autonomy. You will own the architecture of the edge compute platform on the machines, from the NVIDIA Thor-class module and its I/O up through the Linux BSP, drivers and middleware, to the services that perception, data collection and autonomy run on. The platform is being redesigned as the fleet grows to 500+ machines and beyond.
It handles multiple cameras, LiDARs, radars, GNSS and IMUs at terabyte scale per machine, in dusty, high-vibration conditions. This is a hands‑on edge computing role: it needs direct experience with NVIDIA edge hardware and its software stack, not general cloud or IoT architecture.
- Define the edge hardware architecture around NVIDIA Thor-class modules (Jetson AGX Thor and IGX Thor): sensor I/O, camera and network interfaces, storage, power and enclosure.
- Own the low-level software stack: NVIDIA Linux BSP, kernel and drivers, device tree, boot and over-the-air updates.
- Design the middleware layer on ROS 2 (Jazzy) and NVIDIA Isaac ROS, including the migration to ROS 2.
- Set up precise time sync across all sensors (GNSS/UTC, PPS, PTP) and high-throughput logging that doesn't drop data.
- Give perception, data reduction and autonomy services a common driver and service layer, so sensors can be added or removed without redesign.
- Make the platform run onboard GPU inference (detection, segmentation, fusion) and event-triggered capture within compute and thermal limits.
- Build in fleet operations: remote enable and disable, health and diagnostics reporting, and secure, encrypted data handling.
- Decide with the robotics team whether data logging and autonomy compute share hardware or stay separate.
- 8+ years in embedded or edge systems for robotics, autonomous vehicles or industrial machines.
- Direct, hands‑on experience with NVIDIA Jetson or IGX platforms, ideally the Thor generation.
- Hardware design and I/O integration: camera interfaces (GMSL, MIPI CSI), automotive Ethernet, CAN, and storage throughput.
- Linux BSP, kernel and NVIDIA driver development and customization (Jet Pack, L4T).
- The NVIDIA software ecosystem: CUDA, TensorRT, Deep Stream and Isaac ROS.
- ROS 2 at production quality: nodes, DDS tuning, rosbag2 or MCAP logging, and performance under high sensor load.
- Time sync and multi-sensor integration for cameras, LiDAR, radar and GNSS/IMU.
- Ruggedized deployments: off-road, construction, mining or agriculture machines.
- Sensor calibration and online monitoring of calibration health.
- Edge data reduction, compression and encryption for large sensor fleets.
- Fleet device management, OTA and remote diagnostics at hundreds to thousands of units.
- Safety-related embedded development (ISO 26262 or ISO 13849).
Skills:
NVIDIA Jetson / IGX (Thor), Jet Pack / L4T, Linux BSP and kernel drivers, CUDA, TensorRT, Deep Stream, Isaac ROS, ROS 2 (Jazzy), DDS, GMSL / MIPI CSI, CAN, PTP / PPS time sync, C++.
Proactive Global is committed to equality in the workplace and is an equal opportunity employer.
(If this job is in fact in your jurisdiction, then you may be using a Proxy or VPN to access this site, and to progress further, you should change your connectivity to another mobile device or PC).
View / Apply for Jobs
Matching My Jurisdiction