The ADE robotics project carries out the final field tests
Madrid, 26 April 2021.- After nearly two years of work Bremen in Germany hosted from 18 March to 16 April the final tests of the ADE (Autonomous DEcision Making in very long traverses) space robotics project from 15th of March to 16th of April, rescheduled from the initially-intended venue of Fuerteventura due to COVID-19 travel constraints.
ADE falls within the Strategic Research Cluster (SRC) in space robotics technology, coordinated by the H2020 PERASPERA project. Its remit is to develop and test a planetary rover system with very long traverse capabilities by independently taking the decisions required to progress, reduce risks and seize opportunities of data collection.
Coordinated by GMV and run with the collaboration of 13 partners from all over Europe*, ADE is dedicated to autonomous robotic decision-making, targeting specifically surface robot planetary missions performing very long traverses in unknown environments. ADE draws on the ERGO (European Robotics Goal-Oriented Autonomous Controller) autonomy system developed under GMV leadership in the first SRC phase.
The robotics platform used for field testing the project’s technology was the SherpaTT rover developed and provided by the Robotics Innovation Center of the German Research Center for Artificial Intelligence (DFKI). SherpaTT is a desert veteran that has already carried out simulated space missions in the Utah desert (USA) and Morocco under the GMV-led ERGO project, in 2016 and 2018, respectively.
Preliminary tests were held in November 2020 to February 2021 in a 7m x 7m indoor sandbox at DFKI premises in Bremento validate most of the onboard systems: localization, perception, guidance, manipulation of the robotic arm in motion, FDIR (Fault Detection, Isolation and Recovery), dynamic planning, and science agent. After gaining excellent results there, ADE moved on to final planetary exploration tests in a custom-built terrain comprising diverse obstacles and compacted sand.
For 5 weeks the robotic technology was put through its paces, testing navigation autonomy using the perception and localization cameras, sample-collection and -depositing with the robotic arm, automatic goal-oriented mission planning (travel to a fixed point, carry a sample from one point to another, move on to an image-taking point), plus the possibility of ad hoc scientific target recognition as the opportunity arises.
ADE came through the tests with flying colors, pulling off a long and totally autonomous traverse of almost 500m with the rover SherpaTT in a record time of less than three hours. The system incorporates all technological components developed in the previous SRC phase.
Jorge Ocón, Head of Onboard Autonomy Division-Space Segment & Robotics, GMV and project manager of ADE argues that «these tests have been crucial for validating these robotic applications. As well as addressing future space exploration and exploitation needs they also show potential benefits for other robotics areas back on earth.”
The tests, remotely coordinated by GMV with the participation of all 13 ADE partners and followed by members from the Programme Support Activity (PSA), the European Commission and consortium members, were a resounding success. DFKI, as provider of the robotic platform, took charge of commissioning and local control.
The technology developed under ADE is designed to meet future space exploration rover needs. Its goal-oriented autonomy system, apt for various space robots, can also be deployed on the ground for robots working in harsh environments, e.g., nuclear plants, rescue operations, or oil & gas industry.
(*) ADE partners: GMV as coordinator, GMV NSL (UK), King’s College London, Thales Alenia Space Italy, Airbus Defence and Space Ltd, University of Oxford, UGA-Verimag, Magellium, JOANNEUM RESEARCH, Trasys, German Research Center for Artificial Intelligence (DFKI), University of Malaga, University of Salento, and Politecnico di Bari.
This project has received funding from the European Union’s Horizon 2020 research and innovation programme under grant agreement No 821988