MONTREAL, CANADA—Immervision, a developer of advanced vision systems combining optics, image processing, and sensor fusion technology, has announced it was chosen by Halodi Robotics to provide its next generation vision system for its humanoid service robot platform, capable of handling both human and computer vision.
Robots can perform a wide variety of tasks that are difficult or dangerous for humans to perform. To intelligently and accurately perform these tasks, they require mobility and the aptitude to explore, understand and function within various environments. For a robot to work autonomously or problem solve with human assistance, it requires sensor data to understand the human world and the ability to process the data and respond in a timely manner. The difficulty is that for certain applications, a robot needs to combine both human and machine vision capabilities. While performing independent tasks, machine vision suffices, but when functioning in partnership with a human operator, human vision capabilities are needed, hence, the need for a next generation vision system.
“Autonomous robots are intelligent machines able to perform tasks without human control and constant supervision, designed to off load repetitive tasks, such as robot vacuum cleaners and drones that move freely around their physical space without being physically guided/controlled by a human,” says Patrice Roulet Fontani, VP Technology, Immervision. “However, no situation is ideal and invariably algorithms controlling robots and a reliable foundation for their AI learning processes enables the robot’s autonomy. Robotic autonomy is a consequence of several factors. For this to be achieved, high-quality input data and accurate perception of the environment are key factors. “Immervision and Halodi Robotics are working together to integrate a solution that will provide a path forward to make this robotics vision system a reality.”
Fontani explains that the vision system is based on ultra-wide-angle panomorph lens design that covers up to a 360° FoV. It includes the human comparable VFoV feature that is key for humanoid robots. “The Halodi Robotics vision system is built using the latest high-quality image sensors that maximize the light capture,” he says. “Such sensors are also capable of processing the acquired data much faster than previous versions, resulting in an extremely agile vision pipeline.” He adds that the large FoV allows Halodi robots to more effectively detect situations, offering maximized situational awareness.
Immervision’s next-gen vision system enables machines with both visual systems, for multiple sensors and wiring. Robots can now navigate within hospitals or dangerous workplaces, and the vision system can be customized to the environment and specific tasks. The vision system will enable Halodi Robotics robots to better perceive their environment, collect better quality vision data for AI/ML pipelines, as well as providing an immersive VR experience to the operator of the robots.
Fontani says that there is no single camera today that can match the human eyes in terms of adaptability, resolution, sensitivity, and dynamic range. Even if a machine deploys a wide variety of sensors (camera, radar, lidar/ToF, sonar, thermal, microphone, etc.), which can outperform human capabilities in many cases, there are still limitations. “For example, currently, a camera cannot achieve human eye performance at the same speed a human can make sense of imagery input,” he says. “Human vision is very efficient and, in combination with human brain power, it can convey a meaning to an image input in fractions of seconds, while a vision system needs to correlate information from image input, auxiliary sensors, and AI algorithms to do the same job. Also, when you are dealing with multiple sensors it can lead to contradictory conclusions.”
Fontani says that is the problem Immervision is tackling. “We are inspired by biological vision and we work to design vision systems to outperform human vision,” he says. For example in 2004, the company introduced to multiple industries bioinspired freeform lens designs combining super wide-angle FoV with augmented resolution. “We have spent the last two decades working with the state-of-the-art optical technologies such has meta-lens and active-lens, always pushing the boundary of machine sensing and perception. This better-quality data and efficient algorithms are key to allowing robots to perceive their environment, a requirement to enable autonomy.”
“Robots have incredible potential to expand human capabilities and performance in the real world,” says Nicholas Nadeau, CTO of Halodi Robotics. “Immervision is able to help our EVE service robots create even greater value, by allowing human operators, via a virtual reality (VR) headset, to respond with human intuition capabilities to the anomalies that may arise as robots interact with their dynamic environment.”
About Immervision
With over 20 years of innovation, Immervision creates solutions that see beyond human vision. Its Deep Seeing technology and renowned experts in wide-angle optical design and image processing enable smart devices with superhuman eyes to capture high quality visual and contextual data. The company invents, customises, and licenses wide-angle lenses and imaging software technology for AI, machine vision and user applications, from capture to display, in the mobile, automotive, robotics, security, and other industrial and consumer product industries. For more information, visit: www.immervision.com.
About Halodi Robotics
Halodi Robotics builds service robots to extend growing workforces in security, retail, logistics, and healthcare. Its roboticists design and develop affordable humanoid robots with the skill to operate in unstructured environments where people and machines must work safely together. Operating in near silence with human strength, its robots enable independent automation for everyday tasks. Halodi Robotics bases its global teams at its Norway headquarters with satellite offices in Canada, Italy, and the U.S. For more information, visit www.halodi.com.
