ASI receives Phase Two funding for deep learning multi-sensor fusion development

Autonomous Solutions, Inc. (ASI) recently announced that it has been awarded a Phase Two grant from the US Army Combat Capabilities Development Command Ground Vehicles Systems Center (formerly TARDEC). Based on the progress achieved during Phase One, ASI was chosen to continue development of a Deep Learning (DL) architecture that will support sensor fusion in environments with limited, or no, GPS.

Specifically, ASI is making rapid advancements in triangulating data inputs from traditional cameras, LiDAR, and radar to feed machine learning that will provide clearer visibility, predictability, and safety in environments where GPS integrity is restricted or where GPS cannot be utilised at all.

“The objective is to create clearer real-time understanding of an autonomous vehicle’s surroundings, especially when navigating through compromised weather, environments, or conditions,” said Jeff Ferrin, Chief Technology Officer at ASI. “As self-driving vehicles advance, especially for industrial use, the need to utilise machine learning, deep learning, and other artificial intelligence algorithms to improve performance in challenging environments only increases. Therefore, the success of this project is critically important – not only for the direct application within the US military, but for applications across ASI’s multiple lines of business.”

In the case of a deep learning architecture that fuses information from LiDAR, radar and cameras, the innovation could not come soon enough for some industries – especially mining.

“As global mining operations re-evaluate orebody economics and redesign mines as a result of automation, mining operations will become increasingly complex and dependent on technology. By association, the need for advanced visibility and situational awareness increases exponentially,” explains Chris Soccio, General Manager of the Ferrexpo Yeristovo operations. “In locations where GPS or communications networks are compromised or unreliable, the ability to leverage machine learning fed by three diverse input methods becomes not only immediately desirable, but essential to ensure system redundancy for safe and efficient mining.”

ASI expects to complete the Phase Two assignment by September 2022.