While most robots in space are still passive bystanders or in any event controlled from Earth by humans, thanks to the DFKI they could soon be operating independently and over extended periods in the extreme conditions of outer space. The DFKI's Robotics Innovation Center is currently developing pioneering hardware and software concepts and putting them through their paces during 'analog missions' here on Earth. The scientists involved in these projects are now showcasing their research findings and robotic space systems at the DFKI stand in Hall 2.
Given that piloting a robot remotely from Earth is impractical, particularly due to the delay in communication with far-away celestial bodies, it is hoped that space robots in the future will work independently. Led by Prof. Frank Kirchner, the team at the DFKI Robotics Innovation Center in Bremen is therefore developing autonomous robot technologies for use in outer space that are designed to be aware of everything going on around them thanks to numerous different sensors. Artificial intelligence methods and algorithms, such as machine learning processes, are set to enable robots to act and make decisions independently - and even learn from their behavior. After all, as the DFKI researchers suggest, this is the only way to ensure they can be deployed for lengthy planetary and orbital missions without human intervention.
But this doesn't mean astronauts will be out of a job, as they will be able to work hand-in-hand with the robots in outer space - to build infrastructure, for example. The DFKI is in Hannover to demonstrate how it is using different levels of autonomy in this context, with its robots designed to act either more or less autonomously depending on the complexity of the task at hand. If a robot gets stuck on a problem, the astronaut’s job is to step in and teach it a new behavior pattern. What's more, the DFKI is looking into new procedures for analyzing and identifying intention to ensure successful human-machine collaboration. Using these procedures in conjunction with physiological data, for example, the robots should be able to identify their human colleague's emotions and state of mind and factor this in when planning and improving their activities.