5G industry news Technologies | Devices & Sensors

5G for Manufacturing: How 5G is enabling autonomous last-mile logistics

  • 9 minute read
  • Published by Crispin Moller on 21 Jun 2022
  • Last modified 17 Jun 2022
A pioneering live trial is currently underway at the Nissan factory in Sunderland, testing 5G’s ability to boost productivity through the use of autonomous trucks to move parts and assemblies across the plant. In the fifth of his series, Jonny Williamson explores this world-first innovation.

Manufacturing is a carefully choreographed dance that requires many multiples of people and parts to precisely hit their mark at exactly the right time. It’s an analogy that applies to most mass-volume production but is especially apt for how cars are made. 

An automotive assembly line produces hundreds of finished cars a day, each containing tens of thousands of parts. The supply of these parts is in perfect synch with the production schedule. Stock is replenished ‘just in time’. Any later would interrupt production; any earlier would require buffer storage areas and goods to be double-handled (moved and then retrieved).

This impressive feat is made easier by key suppliers locating their operations nearby, some even occupy a building on-site. Although convenient, this results in numerous journeys being made every day, often over very short distances, between nearside suppliers and the factory.   

The Nissan Motor Manufacturing plant in Sunderland is a prime example. The largest car factory in the UK is adjacent to the UK Nissan Distribution Centre and has several on-site suppliers. 

Nissan Factory

Nissan’s Sunderland manufacturing plant is the size of 50 football pitches and two cars roll off its production line every minute. Since opening in 1986, the site has produced more than 10 million cars. 

Pic: Nissan Motor Manufacturing UK

Several thousand HGV movements a day service the plant. Some from very close by, some from much further afield. All arriving at drop-off points strategically located along Nissan’s 7km long production line. In short, it provides an ideal environment to test whether an autonomous capability can boost productivity and efficiencies, safely and securely. 


The ground-breaking 5G-enabled Connected Automotive Logistics (CAL) project is doing exactly that. The £4.9m operation, part-funded by the Department for Digital, Culture, Media and Sport’s £200m 5G Testbeds and Trials Programme, is investigating the practicalities of taking remote control (teleoperation) of an autonomous electric HGV in a live industrial setting.

5G-CAL project partners and expertise:

  • North East Automotive Alliance: Project lead 
  • Sunderland City Council: Project management and regional coordination 
  • Newcastle University: Research and dissemination 
  • StreetDrone: CAV provision and build 
  • Vantec: Logistics expertise 
  • Nissan: Site coordination and operational needs 
  • Coventry University: CAV cybersecurity 
  • Perform Green: Technical and quality assurance, collaboration lead 
  • Connected Places Catapult: Impact assessment, evaluation and dissemination

Less monotony, greater productivity

Autonomous driving is frequently in the headlines, driven in large part by Tesla and its controversial CEO and co-founder. This level of autonomy typically has a driver present. If the vehicle stops for whatever reason, someone is physically there to take control and continue the journey, either to completion or until safe to hand back control. 


This level of autonomous capability doesn’t overcome two serious issues affecting the logistics industry – a shortage of HGV drivers (in the region of 70,000) and spiralling wage bills. It also doesn’t address the monotony of travelling backwards and forwards along the same short route day in and day out. The use case at Nissan Sunderland, for example, is less than 2km.  


By combining teleoperation with autonomous technology, 5G-CAL is helping to overcome all these issues. It’s also helping to improve operational efficiency, driver satisfaction and safety, reduce costs and potentially increase the driver: vehicle ratio from 1:1 to 1: many. 


I sat down with Richard Barrington, principal consultant and head of business development for Perform Green, to discuss the lessons learnt and forward opportunities.

How does your set-up differ from Automated Guided Vehicles, which have become common in factories and distribution centres?


When our vehicle encounters something it doesn’t understand or hasn’t encountered before, it stops and triggers an alarm. In a lot of operating environments, if an AGV stops someone has to physically go and restart it. That doesn’t present much of an issue within a confined space but our use case is a 2km stretch of road. Having to trek out to the vehicle costs time and money. That’s where remote control comes in. 


Our teleoperator rig is an exact replica of the in-vehicle driving station and three screens display a direct camera feed from the left, right and straight ahead of the vehicle. The user experience is the same as if they were sitting in the cab and the sensory information is the same. With a sub-10 millisecond latency, the information is also being presented in what is effectively real-time. 


When the vehicle stops and triggers an alarm, the teleoperator sits down in the rig and makes an informed decision as to how to proceed until they feel comfortable passing control of the vehicle back to the autonomous software.

5CAL Truck

New project partner, Dutch vehicle manufacturer Terberg, has taken one of its tractor units, often seen operating in airports and ports, and retrofitted it to be electrically powered and drive by wire. StreetDrone has installed its autonomous software on that vehicle and it has been fitted with a wide array of sensors, LiDAR and cameras

Pic: Sunderland Smart City

Why is 5G so crucial to 5G-CAL?

Many manufacturing sites have strict speed limits of 5 or 10kmph. In theory, 4G could be used but as soon as you travel beyond the confines of the facility, you’ll want to go a lot faster. Higher speeds mean longer stopping distances and that’s where the ultra-low latency of 5G becomes very important. We need to know we can stop the vehicle on a dime to ensure the whole operation is safe and secure.


Additionally, the HGV has been fitted with multiple cameras and LiDAR units and lots of sensors. These produce lots of data, around 100Mbps, that has to be taken off the vehicle and presented to the teleoperator who, in our case, is sitting in the Vantec building. That's very hard to do with any technology other than 5G. It is also very different to a Tesla, for example, where all the decisions by the autonomous software happen locally.


The reliability of 5G is another major factor. The performance of our private network, using Ofcom Shared Access spectrum n77, is consistent from end to end. You don’t get that with 4G. 5G is also very secure, particularly with a private network. The only devices that can talk to our network are those we’ve authorised and issued sims to.


What stage has the project reached? 

StreetDrone has finished the engineering on the vehicle and it’s at Nissan Sunderland. We're very fortunate that Nissan has not only a private road network but also a skid pan and a test track on-site. 


Currently, we are practising all the basic manoeuvres on the skid pan, with a particular emphasis on the handover between the autonomous software and the teleoperator and back again. Our focus is on making sure that it works under all circumstances and demonstrating that our system is more responsive and therefore safer than having someone sat in the cab.


Tests have been ongoing since April and we expect to progress onto the private road network in June. From there we’ll be looking at two things. The first is scale up. We’ve proved it works with one vehicle, what happens when we have six? How does the network stand up? Do we have one-person teleoperating six vehicles or is 1:6 too high a ratio? 


The second is negotiating a more complicated route but still within a private road network. We can go to another distribution point which means having to go through two sets of traffic lights, a roundabout, security gates and a humpback bridge. The route has traffic from multiple distributors going to and from the plant so it’s a far more real-world operating environment.

Are you able to quantify any of the expected efficiency improvements? 

From what the project has achieved so far, a plant similar in scale to Nissan should be able to see a £2.5m to £3m efficiency saving per annum.


We are also looking at creating efficiencies at the other end of the process. Nissan Sunderland produces around 3,000 cars a day, 80% of which are exported through the Port of Tyne. That means hundreds of car transporters carrying thousands of cars, not to mention the containers full of parts arriving at the port each day. 


Once we have proven our system works across multiple vehicles and that it can work across a much more complicated road network, which will likely take 24 – 36 months, the next step could be overseeing transport between Nissan and the port. 


That would likely involve having to move between private networks and potentially the public network. How would those handovers work? Would the private network provide our required performance characteristics? Your standard MNO network is designed to deliver data down to devices not for bringing data up from them.

Does your system have applications beyond automotive and other large discrete manufacturing plants?

We’ve been working with Connected Places Catapult and Newcastle University to explore the potential for 5G enabled autonomous vehicles to be deployed more widely in other industrial and social settings.


Other use cases include airports, ports, public transport services, and campuses. Most supermarkets, for example, operate in much the same way that a car plant does with hundreds of trailers getting dropped off and picked up every day. 


There are significant costs associated with delaying an aircraft, much like with an assembly line. Holding up a flight because the baggage isn’t at the right place at the right time does more than frustrate passengers. These areas all present massive opportunities to become more efficient and more use cases will emerge as we continue to refine this technology.

Related projects