A new development in affordable, open-source mobile networks which enables near-real-time control of robotic arms could help doctors work on patients in remote locations in the years to come.
Researchers from the University of Glasgow’s James Watt School of Engineering have used cheap off-the-shelf hardware to build a 4G LTE mobile network which connects a haptic controller to a robot arm. The system allows users to direct the arm’s movements with very low latency, enabling a high level of control.
In a new paper published in the journal Communications Engineering, they demonstrate how they used the system to perform mock dental exams on a pair of dentures, highlighting its potential for use in medical procedures in the future.
The team’s system is built using the Open Radio Access Network (O-RAN) framework, which uses open-source software to control mobile network hardware.
They repurposed a USB network dongle, more commonly used for consumer mobile internet, to create stable connections between the haptic input device, the robotic arm, and a computer configured to act as an intelligent base station.
Using specialised software called xApps, they could monitor and fine-tune the performance of the system’s signal quality, data rates and latency.
In their lab tests, the team were able to enable communications between the base station, the controller and the robotic arm with a bandwidth of 10Mbps. The high-quality connection allowed them to control the arm to simulate a dental exam on dentures with less than a second of latency and minimal signal loss.
The system’s mobile dongle also helped the team create a network which drew considerably less power than comparable connections using software-defined radio (SDRs), which are more commonly used in similar tasks. Their low-budget system used just 4,5 watts, a 90% reduction on the 45 watts required by traditional SDRs to perform the same activities.
Dr Saber Hassouna of the James Watt School of Engineering and the paper’s first author, says: “The O-RAN framework holds a great deal of potential for enabling intelligent, data-driven, programmable and virtualised networks, but a significant amount of work remains to be done to demonstrate that potential being achieved in the real world, beyond theoretical modelling.
“The testbed we’ve developed here using commercially-available hardware shows that O-RAN can be used to enable excellent performance in robotic teleoperation, which is a complex task. For applications like dental procedures, the robotic arm must move very smoothly, which requires high data throughput and low latency, both of which we’ve been able to achieve for the first time with O-RAN.”
Professor Qammer Abbasi, head of the University of Glasgow’s communications, sensing and imaging hub, adds: “This is a very encouraging demonstration of the potential of O-RAN to enable fine-grained, close to real-time control of a robotic arm.
“The paper showcases the performance we’ve been able to deliver in a single room with a direct line of sight between the base station and the arm, and we’re currently working on developing the system further to ensure it can deliver the same level of performance at greater distances.
“Ultimately, this could be a step towards creating reliable, affordable methods of performing complex tasks remotely, opening up new applications in medicine, automation, industry and beyond.”
Professor Muhammad Imran, head of the James Watt School of Engineering and the paper’s lead author, comments: “The Internet of Skills is now one step closer with the advent of low-latency, high-reliability communication links. With Open RAN (O-RAN) making these technologies more affordable and accessible, I’m proud that our team is helping to lay the foundation for truly inclusive digital accessibility for all.”
The team’s paper, titled ‘Development of Open Radio Access Networks (O-RAN) for Real-time Robotic Teleoperation’, is published in Communications Engineering, a Springer Nature journal. Researchers from the Indian Institute of Technology Bhilai and the Indian Institute of Technology Delhi co-authored the paper.