Miles Jamison 
- Riverside Research will integrate Tera AI’s DeepSLAM technology into its capabilities
- DeepSLAM uses passive cameras to create 3D maps without relying on GPS
- The technology can support aircraft and unmanned systems operating in disrupted environments
Riverside Research has partnered with Tera AI, a startup offering advanced real-time spatial reasoning for autonomous robotics, to provide mapping and positioning capabilities for air and surface operations in GPS-denied environments.
What Will the Partnership Deliver?
Tera AI’s DeepSLAM technology builds 3D maps of unmapped environments using standalone passive cameras, Riverside Research said Tuesday. The technology is designed to provide location information when GPS signals are jammed or spoofed, helping address gaps that can affect autonomous systems.
Riverside Research will integrate DeepSLAM into its existing capabilities to support positioning in contested and disrupted environments. The organizations aim to close data gaps that affect safety products and intelligence missions.
The partnership builds on Riverside Research’s recent collaborations with technology companies to develop and transition specialized capabilities for defense and intelligence customers.
In May, the nonprofit organization partnered with Plexus to develop reusable modular hardware and software for intelligence and defense customers. The companies used rapid prototyping to move their first joint effort, the Aegis R52L Enhanced Retransmission Device, from concept and requirements identification through prototyping and manufacturing planning in less than nine months.
What Applications Could the Technology Support?
According to Riverside Research, DeepSLAM could support field operations as well as platforms that depend on GPS, including commercial aircraft and global shipping networks.
The technology can also support missions involving battlefield intelligence and target identification.
Riverside Research has also worked with defense technology companies to transition research capabilities into operational systems for national security missions. In 2025, the organization partnered with Anduril Industries to apply formal methods technology to cyber-harden critical defense capabilities at the tactical edge.
Tony Zhang, founder of Tera AI, said AI foundation models, including world models, can drive spatial intelligence capabilities such as visual navigation and real-time mapping.
“As GPS and comms jamming becomes standard, autonomous systems will require real-time spatial intelligence to realize its location, while adapting to changing platform needs.”
Carl Nerup, vice president of corporate development at Riverside Research, said DeepSLAM can provide location information for aircraft and unmanned vehicles when GPS is unavailable.
