Seeing Machines has launched a Physical AI Platform designed to give humanoid robots and industrial automation systems greater awareness of people and their surroundings, with manufacturing among the industries identified as a potential application.
In a media release, the Canberra-based computer vision company said the platform extends its Human-Centred AI approach, developed through more than 25 years of research into human factors and artificial intelligence.
Seeing Machines said its driver and occupant monitoring technology is already deployed in more than eight million vehicles globally.
According to Seeing Machines, the new platform creates a dynamic three-dimensional perception map of people, objects and their surroundings. It is designed to help robots understand spatial relationships, interpret human behaviour, anticipate risks and make decisions in real time.
Seeing Machines CEO Paul McGlone said the technology represented an extension of the company’s work in human-machine interaction.
“For more than two decades, we’ve been teaching machines to understand people,” McGlone said. “With the launch of our Physical AI Platform we are extending that capability into robotics, enabling machines to interact with people safely, naturally and intuitively.”
Seeing Machines said the platform differs from AI systems focused primarily on identifying individual objects or performing predefined tasks by interpreting people, objects and the surrounding environment as a unified scene.
McGlone said this capability would become increasingly relevant as robots move into environments where they work alongside people.
“As robots begin moving beyond research labs and into factories, workplaces, hospitals, homes and public spaces, understanding people and the environment they’re in becomes just as important as understanding the task they’re performing,” he said.
Seeing Machines identified manufacturing, logistics, healthcare, aged care, warehousing, mining and industrial automation as potential application areas for the platform.
