Off-Road Autonomy solutions
ARA provides uncrewed, ground vehicle autonomy solutions for off-road environments. Off-road autonomy presents a set of unique challenges when compared to on-road applications, including dense, cluttered terrain, as well as a lack of GPS and detailed maps. ARA takes a modular approach to these challenges, aiming to provide a full-stack solution that is easily integrated on a variety of vehicles and platforms.
Key Features
Modular/Quick Integration for Drive-by-Wire Platforms and Perception Sensors
Integrated Specific Vehicle Capability Planning
Vehicle and Environment Specific Planning
Optimized for Cluttered, Dense Environments
Minimal GPS Dependency
Optimized for Cluttered, Dense Environments
Vehicle and Environment Specific Planning
Integrated Specific Vehicle Capability Planning
Modular/Quick Integration for Drive-by-Wire Platforms and Perception Sensors
Advanced Capabilities
Individual Wheel
Placement
ARA’s advanced software takes into account individual wheel placement and factors in vehicle underbody clearance, roll, and pitch, enabling navigation in extremely uneven terrain without high-center or tip-over risk.
Self-Supervised
Mobility Learning
Our software uses self-supervised mobility learning to determine the physical characteristics of a vehicle to optimize performance and ease installation onto new platforms.
Unsupervised World
Model Adaptation
Our software uses self-supervised mobility learning to determine the physical characteristics of a vehicle to optimize performance and ease installation onto new platforms.
Advanced Capabilities
Individual Wheel Placement
ARA’s advanced software takes into account individual wheel placement and factors in vehicle underbody clearance, roll, and pitch, enabling navigation in extremely uneven terrain without high-center or tip-over risk.
Self-Supervised Mobility Learning
Our software uses self-supervised mobility learning to determine the physical characteristics of a vehicle to optimize performance and ease installation onto new platforms.
Unsupervised World Model Adaptation
Our software employs unsupervised world model adaptation to learn interpretation of world model data (confidence, error rate, probabilistic terrain traversability estimates.)
Vehicle and Sensor Agnostic
