C++ and ROS autonomy software linked LiDAR, GPS and IMU data to localisation, route planning and vehicle control on an electric mower tested outdoors.
The autonomous mower during outdoor field testing.
A working autonomous mower and a reusable foundation for other service vehicles.
Why we built it
We used the mower to test the complete autonomy stack.
Boundary wire and random coverage were not enough for the product Elite Robotics wanted to build. The mower needed to map outdoor terrain, detect obstacles, estimate its position and plan complete coverage using practical onboard hardware.
Outdoor testing exposed problems that bench tests could not. The team could see whether a failure came from sensing, planning or vehicle control and work on the right layer.
We separated the compute, sensor interfaces, localisation, planning and control software. That kept the mower practical while leaving the core software ready for other outdoor service vehicles.
How the system fit together
The vehicle and drivetrain already existed. Elite Robotics built the sensing, compute, navigation and control layers around them.
Machine
Electric vehicle platform and drivetrain.
Sensing
GPS, IMU and LiDAR data with aligned frames and timing.
Embedded
Raspberry Pi and NVIDIA Jetson hardware interfaces.
Compute
C++ ROS nodes, message contracts and launch configuration.
Navigation
Localisation, obstacle tracking and coverage planning.
Control
Motion commands passed into the vehicle controls.
Role and approach
Sahil's role
At Elite Robotics, Sahil worked as co-founder, CEO and embedded engineer. He led the product and commercial direction, designed the embedded architecture, selected the onboard compute and integrated the complete vehicle. The wider team built the probabilistic mapping and tracking algorithm. Sahil connected it to the sensors, hardware and controls used in field testing.
How we worked
Used one mower as the proving ground for the complete autonomy stack.
Kept sensing, localisation, planning and vehicle control separate so each part could be tested and replaced.
Brought LiDAR, GPS and IMU data into ROS and made timing and coordinate frames visible before sending motion commands.
Moved from bench tests to outdoor runs and public demonstrations.
What we built
Onboard vehicle software
Raspberry Pi and NVIDIA Jetson integration
ROS nodes, message contracts and launch configuration
Hardware interfaces for power, sensing and motion
Localisation and navigation
LiDAR, GPS and IMU acquisition
Coordinate frame and timing alignment
Probabilistic sensor fusion and multi-target tracking
Outdoor field testing
Coverage and obstacle-aware path planning
Command interfaces into vehicle control
Outdoor testing, demonstrations and tuning
Watch it run
Navigation and sensor fusion
The demo shows range-sensor data being filtered into terrain and object clusters while the system estimates the vehicle path without a prior map.
Autonomous mower field test
The upgraded Elite Robotics mower runs the autonomy software while cutting grass outdoors.
Field testing and demonstrations
The vehicle platform beneath the final bodywork.
Software and vehicle testing in the operating environment.
Presenting the product and target operating environments.
Evidence
Outdoor field tests
Autonomous operation
The platform progressed from bench integration to an autonomous mower operating outdoors.
Complete autonomy stack
System integration
LiDAR, GPS and IMU inputs fed localisation, planning and vehicle control.
NSW MVP grant
Grant support
Elite Robotics received an NSW Government Minimum Viable Product grant.
Slingshot Accelerator
Accelerator program
The company completed the Slingshot Accelerator and received backing through its venture fund.
