Robotics & controls
Embedded systems
Autonomous electric vehicle platform
Elite Robotics
From sensors to motion
A C++ and ROS platform spanning onboard compute, positioning, perception and motion control for an autonomous electric vehicle.
01
The system problem
Challenge
The vehicle needed a coherent software platform across heterogeneous compute and sensors. GPS, IMU and LiDAR data had to be acquired, interpreted and converted into dependable motion while the team was still evolving the physical platform and product direction.
Our responsibility
SSH Tech worked across the autonomy stack, integrating Raspberry Pi and Jetson compute, establishing ROS-based interfaces, bringing sensor data into a common frame, and connecting navigation decisions to vehicle motion.
How we approached it
Defined clear software boundaries between sensing, localisation, planning and actuation.
Integrated each sensor independently before combining data into navigation behaviour.
Used ROS messaging and observable runtime states to keep subsystems testable as the platform evolved.
Balanced prototype speed with interfaces that the growing engineering team could continue to extend.
02
What we delivered
01
Establish the vehicle platform
Created the C++ and ROS foundation across the available onboard compute and vehicle interfaces.
- Raspberry Pi and Jetson integration
- ROS nodes, messages and launch configuration
- Hardware abstraction for evolving components
02
Make the vehicle aware
Brought positioning and perception signals into the same operational picture for navigation.
- GPS and IMU integration
- LiDAR acquisition and processing
- Coordinate-frame and timing alignment
03
Connect perception to motion
Implemented the path from interpreted sensor state through planning to controlled vehicle movement.
- Navigation and motion-planning logic
- Command interfaces to vehicle control
- Field testing and iterative tuning
03
Outcomes and evidence
End-to-end
Autonomy stack
The delivered platform connected GPS, IMU and LiDAR inputs through to navigation and control.
Two compute classes
Onboard integration
The platform operated across Raspberry Pi and NVIDIA Jetson hardware.
Technical stack
C++
ROS
Raspberry Pi
NVIDIA Jetson
GPS
IMU
LiDAR
Capabilities applied
Robotics software
Sensor integration
Motion planning
Embedded compute
Continue exploring
Battery systems and manufacturing automation
2 months to 3 days