Inspired by nature
Over the course of evolution, insects have developed complex skills for navigation, collision avoidance and communication. After ten years of research, Opteran has succeeded in deciphering how the insect brain works and reproducing it in the robot autonomy software Opteran Mind. Instead of using pre-recorded maps, as in conventional localisation methods, the system uses mechanisms from biology and computational neuroscience for localisation. Opteran calls this approach ‘natural intelligence’. Our system is extremely small, lightweight and energy efficient,’ says David Rajan, CEO and co-founder of Opteran Technologies. „And unlike other machine learning approaches, we don’t need data to control a robot.“ Instead, Opteran Mind is purely visual software, with no control centre, network or training in the background. The solution also does not require any additional infrastructure such as magnetic tracks, QR codes or reflectors, which speeds up the start of the project.
SAFELOG plans with new localisation technology
SAFELOG will integrate the localisation software into its mobile robots from 2025. The partnership with Opteran includes further optimisation of Opteran Mind based on the experience gained in the field. Opteran’s approach redefines the localisation of mobile robots,” says SAFELOG Managing Director Mathias Behounek. „Instead of collecting more and more data, the solution generates almost no data. And absolutely basic technologies are enough to run the system stably. That’s impressive.“ While other visual SLAM solutions with cameras cost up to $5,000 or LiDAR SLAMs are available for around $3,000, Opteran requires a CPU in the sub-$20 price range and inexpensive cameras for less than $30. Mathias Behounek is certain: “Opteran fits perfectly into our lean robotics approach, where we install as little complex technology in our mobile robots as the task allows.“