Can 5G deliver real-time cloud control for industrial robots?

Researchers at the University of Auckland, with support from Spark and Nokia, have been investigating whether industrial robots can be controlled in the cloud, via a mobile network. 

Testing the 5G network across five countries 

“The 5G network is a huge leap forward for the commercial world”, says Dr. Yuqian Lu, Lead Researcher and Senior Lecturer in Mechanical and Mechatronics Engineering at the University of Auckland. He has been working to quantify and compare performance between public and private 4G and 5G networks.

Lu’s team has conducted three rounds of testing over six months, sending data between Auckland, Sydney, London, Singapore and Oregon under various network settings to test whether industrial robots could be controlled remotely from a central cloud-based platform in real time. 

“It’s all about understanding the performance of 5G, any limitations, and how we can make it work for us,” says Lu. “Based on our tests, 5G is not yet able to plug and play for all commercial purposes, especially for real-time applications – you will need some expertise to support it. But we can say that in the future, industrial control will be moved to the cloud. It’s fantastic to see the promising improvements in 5G, although it’s not enough yet.”

Support from expert partners and Government funding

The devices and private network were supplied by Nokia. Spark provided access to public network and cellular network expertise and The Science for Technological Innovation National Science Challenge provided the funding, so it’s been a real team effort.

The research is a perfect fit for Spark, says Joseph Wong, Innovation Lead, because it dovetails with the company’s goal to help New Zealanders win big in a digital world. Leveraging 5G will help drive transformation across many industries. 

“Yuqian’s project is close to our heart,” he says. “The future will involve a lot of robotics, industrial AI and automation. His team has already proven themselves and we believe working closely with them can bring a lot of good applications to reality – we see particular benefits for automation in the agriculture and ports. This is contributing to how we think about the future.”

The future of connected robotic devices 

Having completed the first stage of testing, the next step has been developing algorithms to compensate for the limitations of 5G, so that we achieve the required speed and reliability performance. Lu says early iterations are “working nicely”, and he is building a true cloud-based robot at his lab – Laboratory for Industry 4.0 Smart Manufacturing Systems at the Faculty of Engineering.

“We see in long-term, with our technology, a surgeon in Auckland could be performing an operation on a patient in Invercargill, or a fleet of manufacturing robots in New Zealand could be controlled seamlessly from the US, leveraging the true global expertise”, says Lu.“

Although early results show that the network isn’t quite ready for full industrial cloud automation, Wong says the research is helping Spark think about the next steps toward a more connected future. 

“Some of our work around Multi-access Edge Compute and network slicing, for instance, helps us to bring the cloud from the internet to as close as your individual factory, so you could in future deploy the cloud at your premises to run your automation over a dedicated network,” Wong says. “This gives us a north star to say, ‘This is what we need to do next’. It’s very exciting.” 

Related articles