%20(2).png)
Researchers at the Autonomous Systems and Machine Intelligence Lab (ASMIL) at City, University of London study the motion of robots.
Nabil Aouf, Professor of Robotics and Autonomous Systems, and members of his team, including Postdoctoral Research Fellows Duarte Rondao, Zakaria Chekakta, and Abdelhafid Zenati, and PhD students Mahmoud Abdulsalam and Maxwell Hogan, focus on aerospace-based applications and cooperative path planning. They have used an OptiTrack system for tracking for almost a decade.
The issue Their existing lab provision lacked the size and precision they now require. As an example, the team apply algorithms to robots in order that they can sense the world around them and navigate through it. A lack of physical space was hampering their ability to expand the scope of their projects - they were using smaller scale models rather than recreating scenarios that were akin to the real world.
The solution The creation of a new, large scale drone lab provided an upgrade of size and capability. Target3D designed and installed the lab using six Optitrack PrimeX 22s on a custom-built 13x9x5m truss. The upgraded cameras provide increased precision and FOV, and are enhanced by the addition of VR kits, providing capability to add a human into their research projects.
The use This state-of-the-art arena is set to be utilised for two main areas of study:
1. Aerospace-based applications
Lidar-based navigation for landing operations for lunar terrains or asteroid terrains in collaboration with European Space Agency. Mounting a drone platform with tracking sensors and simulating its landing on specific terrain. The performance of the sensors and algorithms will be compared with the ground truth provided by Optitrack.
For rendezvous, a complex space operation bringing two spacecraft into close proximity. One proposed use is for debris removal, to address the issue of space junk orbiting the earth by deploying another satellite to rendezvous and to capture the target of junk, bringing it down to the atmosphere so it can burn up. This requires very high precision in terms of algorithms and the new arena will allow the testing of bigger, more complex targets while keeping the precision from Optitrack.
Drones for precision agriculture. Sending out ground robots to grid-search farm fields for weeds is highly inefficient. Instead, drone platforms with mounted cameras will be investigated for use as scouts to precisely detect these undesired plants from above and relay their position to the ground agent, who can then remove them. Optitrack will be used to compare the real position of the targets against the one estimated by the algorithms.
Explainable artificial intelligence (AI). Commercial applications of AI for unmanned aerial vehicles have grown in recent years, but their black-box nature makes them somewhat unpredictable. The team's goal is to develop AI algorithms for object recognition and scene understanding that can be better understood by a human developer. Drones will carry out complex tasks, pre-programmed using the indoor localisation information provided by the Optitrack system. This can then be used to gather data for training and testing the algorithms.
2. Cooperative path planning
Using two or more robots to plan an activity in parallel in the most efficient way. The team work with custom-built drones and robots, including The Jackal, a four-wheel all terrain compact robot, with their aim being to make both platforms work cooperatively with obstacle avoidance. With Optitrack in the arena they now have the ability to track the robots in real time with two possibilities; either to use the information from Optitrack in a feedback loop - like an indoor GPS - to localise the robots, or to test their own onboard algorithms by recording the data and using it to compare the performance.
The feedback
“This is an amazing facility which gives our ASMIL researchers at City the chance to conduct world-class research projects with our partners. It will also help to deliver high quality demonstrations to engineering undergraduate and postgraduate students.” Prof Nabil Aouf
Duarte Rondao noted it was refreshing to see how the tech has evolved and improved over the last decade. On working with the Target3D team, he says:
“Definitely 5 stars. Target3D have more than accommodating with us. The arena took many meetings and discussions to coincide with the various teams, and the experience has been 100% positive. We have a great working relationship and we will stay in touch for future endeavours.”
The Autonomous Systems and Machine Intelligence Lab is based in the Department of Electrical and Electronic Engineering within the School of Mathematics, Computer Science and Engineering at City, University of London.
For robotics, tracking and motion capture solutions, speak to Target3D. View more about our work with universities here.
Comments