One of the exciting things about 5G is that it could open up all sorts of new use cases for mobile data, rather than just making current things faster, and one example of that is smart farming, with a new 5G testbed currently looking at how the tech could be used to monitor fields.
Part of the 5G Rural Integrated Testbed (5GRIT) project, the trial will see Kingston University’s Robot Vision team exploring how drones could use 5G technology to fly over fields of crops and livestock to monitor them.
Why does it need 5G? Well, using current mobile data you would need to handle most of the processing on the drone itself, which would require hefty hardware and make it impractical to fly for extended periods due to the weight.
But with 5G the speed will be high enough and latency low enough that the processing can all be handled externally in a data centre elsewhere on the farm or even offsite, leaving the drone lightweight and potentially able to fly for hours at a time.
Deep learning, powered by 5G
And there’s a lot of tech involved. The trial is using drones with onboard visual and infrared cameras – those bits can’t be offloaded, but the tech powering them can, and that tech will allow the drones to tell if an animal is sick, injured or missing, as well as monitoring and counting crops, and checking for signs of weeds and disease.
The drone would be able to detect these things in part using a type of machine learning called deep learning, which would allow the drone to tell the difference between normal behaviour and irregularities, without being specifically programmed to. This assessment all happens in real time and wouldn’t be possible over 4G.
The trial is just one part of the 5GRIT project, which itself is just one of six 5G projects across the UK they have received a share of £25 million from the government – with £2.1 million going to the 5GRIT.
Other recipients include a healthcare testbed in Liverpool, manufacturing and security trials in Worcestershire, plus 5G projects involving tourism, self-driving cars and broadcast.
