Valkyrie is a human sized platform standing about 1.8 meters tall and weighing approximately 125 kilograms. The system features a human like form factor so it can work alongside people and carry out high risk tasks in locations that may be hazardous or inaccessible to human operators.
NASA originally developed Valkyrie as a technology testbed for missions that would send robots to Mars years before human crews. The agency envisaged humanoid systems like Valkyrie performing pre deployment duties and maintaining infrastructure and assets on the Martian surface ahead of astronaut landings.
When the platform arrived in Edinburgh in 2016 it was one of the most advanced humanoid robots in the world. It was also the only Valkyrie unit outside the United States and one of only three prototypes ever built, making its presence in the UK a rare asset for robotics research.
On delivery, Valkyrie could walk on flat terrain and execute basic tasks such as holding and manipulating objects. Over the past decade, researchers at the Edinburgh Centre for Robotics significantly expanded these capabilities using artificial intelligence and machine learning techniques to help the robot interpret and respond to complex environments.
Teams improved Valkyrie's handling and walking performance, enabling more robust motion over uneven or unpredictable surfaces. They also enhanced how the robot fuses information from its onboard sensors so it can better understand its surroundings and move safely in close proximity to people.
The research program focused on whole body manipulation and perception in humanoid robots. Scientists worked on methods that allow Valkyrie to adapt to changing conditions using data driven control, and to link what the robot sees directly to how it plans and executes movements fast enough to operate in real world scenarios.
Work with Valkyrie took place at the Edinburgh Centre for Robotics, a joint initiative between the University of Edinburgh and Heriot Watt University. Dozens of PhD students and researchers used the platform to advance humanoid control, motion planning and perception, helping to train a generation of specialists in the field.
The project received support from the Engineering and Physical Sciences Research Council, part of UK Research and Innovation. This backing helped sustain long term experimentation required to refine learning based planning and control methods for complex humanoid systems.
Reflecting on the program, former University of Edinburgh student and current Touchlab Chief Technical Officer Dr Vladimir Ivan described hosting Valkyrie as a rare opportunity at a time when few humanoid prototypes were available worldwide. He said the robot enabled fundamental advances in mobility and stability research and helped establish Edinburgh as a robotics hub with world class research and a growing ecosystem of robotics businesses.
Humanoid robotics research in Edinburgh will continue using Talos, a 1.75 meter tall platform the University acquired in 2020. Scientists employ Talos to study how robots walk, maintain balance, use tools and apply machine learning to adapt to dynamic, human centric environments through dyadic human interaction.
Advances from the Talos program and related work could support applications in assisted living, healthcare and other domains where robots collaborate closely with people. Researchers aim to translate lessons learned from Valkyrie and Talos into safer, more capable humanoid systems for both space and terrestrial use.
Professor Sethu Vijayakumar, Personal Chair in Robotics and Director of the Edinburgh Centre for Robotics, said it was a bold decision to invest heavily in humanoid research in the 2010s when the scalability of adaptive, learning based planning and control was uncertain. He noted that this work helped shape today's data driven humanoid research wave and that Valkyrie, with its advanced NASA hardware, acted as a trendsetting platform for the field.
Related Links
University of Edinburgh
All about the robots on Earth and beyond!
| Subscribe Free To Our Daily Newsletters |
| Subscribe Free To Our Daily Newsletters |