Ghost Robotics CEO on 'Deployment Gap'
Gavin Kenneally, CEO of Ghost Robotics, remarked on the significant difference between controlled robotics demonstrations and the challenges of operating in live industrial environments. Speaking on the Robot Report Podcast, he noted that while the "deployment gap" is closing, it remains a major hurdle for the industry.
- Ghost Robotics' Vision 60 quadruped robot is designed for "unstructured" environments where wheeled or tracked robots fail, a key factor in addressing the deployment gap in sectors like defense, public safety, and industrial inspection. This capability allows them to navigate difficult terrains like mountains, construction sites, and disaster zones. - The "deployment gap" is a well-recognized chasm between the capabilities of robots shown in controlled lab settings and their actual performance in complex, real-world scenarios. Factors contributing to this gap include regulatory hurdles, the need for more robust and adaptive systems, and the challenge of ensuring safety and reliability in unpredictable environments. - Gavin Kenneally, who holds a Ph.D. in Mechanical Engineering from the University of Pennsylvania, co-founded Ghost Robotics and was responsible for the mechanical design of the Vision 60 before becoming CEO. The company's approach has been to simplify the robot's hardware and push complexity into the software, creating a more responsive and life-like machine. - Legged robots like the Vision 60 are mechanically complex with many active "degrees of freedom" (joints), making control in unstructured environments a significant challenge. Ghost Robotics utilizes the motors themselves to "feel" the ground, creating virtual springs through software to improve balance and responsiveness without mechanical springs. - The Vision 60 is already in use by the U.S. Department of Defense for perimeter security at Air Force bases and is being tested by the Department of Homeland Security for border patrol. These applications leverage the robot's ability to carry various payloads, including thermal cameras and sensors for detecting chemical or nuclear threats, keeping human operators at a safe distance. - Advances in embodied AI and learning-based controllers are critical for closing the deployment gap by enabling robots to learn from interaction and adapt to new situations. This allows them to generalize from training to perform tasks in dynamic, real-world settings like manufacturing and logistics.
