SFL Scientific Demonstrates Chemical Defense Capabilities of Onboard AI

The company’s OptoAI identified visual, thermal, and chemical threats in real-time on board a small UAV at a US DoD field exercise, displaying markers in an augmented reality sandtable for command and control By DA Reporter / 17 Aug 2022
OptoAI_Collage
Follow DA

SFL Scientific has debuted the chemical defense capabilities of its OptoAI Artificial Intelligence (AI) system at the Defense Threat Reduction Agency’s Chemical Biological Operational Analysis 2022 scenario-based exercise. It received a positive evaluation from warfighters for improving mission performance and force protection.

OptoAI’s mission-specific algorithms enhanced existing platforms and sensors, aided by an optional low-Size, Weight and Power (SWaP) computation module. These new capabilities follow OptoAI’s prior success with critical infrastructure inspection for defects and thermal anomalies, and open the doors to diverse military and civilian use cases.

Small unmanned systems hardware is increasingly able to support AI and Machine Learning (ML) algorithms. However, beyond mainstream computer vision and mapping such as tracking people and vehicles, it is rare to see AI-based capability on small Unmanned Aerial Systems (sUAS), let alone on board or in real time. That is the gap AI consultancy SFL Scientific fills, having become a respected provider for custom AI algorithms.

“Demand was so high for mission-specific AI on unmanned systems that we created an entire division dedicated to developing it,” said Mike Segala, PhD, high-energy physicist and CEO of SFL Scientific.

Using specialized sensors, such as those that measure Chemical, Biological, Radiological and Nuclear (CBRN) threats, generally requires intensive training, close proximity to potential hazards, and operator focus. Instead, OptoAI fuses data for robust identification and location, then displays real-time results for intuitive decision support.

Alex Moseson, PhD, Director of Federal Solutions for SFL Scientific, remarked, “It is always an honor to be embedded with warfighters, collaboratively advancing technology.” Transitioning technology from the lab to the real world is challenging, but was part of Dr. Moseson’s prior role at DARPA.

From the video, it seems the OptoAI team found a practical use for the floating sandtable concept usually limited to sci-fi. Markers for objects of interest pop up on a 3D map in real time; using the sandtable and integrated visual and thermal video feeds, the commander directs warfighters away from an ambush.

That kind of situational awareness is dual use – civilians such as first responders, farmers, law enforcement, maritime search & rescue experts, 5G providers, and energy companies have expressed interest too. OptoAI is mission-configurable, meaning that it is agnostic to platforms and sensors, integrates with de facto software and standards like MAVLink, and its core capabilities can be rapidly adapted to unique needs.

Dr. Segala added, “For a nuclear energy client, OptoAI identifies defects in critical infrastructure with an accuracy of 93 – 99%, versus 80% for human inspectors, in a fraction of the time. And it’s not about replacing people with AI; it’s a tool to empower people to get things done more effectively, cheaper, and safer.”

“OptoAI can address many challenges. To maximize reach and impact, we’re always interested in hearing from people representing operators, platforms, and sensors who’d like to explore new applications,” said Moseson.

Posted by DA Reporter Connect & Contact
Latest Articles