5GEM use case: Ford deploys real-time process analysis & control

Critical automotive manufacturing processes, like robotic welding of components for electric cars, require detailed monitoring and quick configuration changes to optimize the quality of the finished products. By analyzing data in real time from 5G-enabled sensors installed throughout the factory, auto manufacturer Ford Motor Co. is able to monitor the production process and make rapid adjustments automatically in response to changes in environment, input materials and other factors. This is one of the first use cases for the 5G Enabled Manufacturing (5GEM) project, a consortium of eight partner organizations (ATS Global, Ford, HSSMI, Lancaster University, TM Forum, TWI, Vacuum Furnace Engineering and Vodafone) exploring use cases for 5G private networks in different manufacturing environments. The project is supported by the UK government’s 5G Testbeds and Trials Program. The 5GEM team is implementing real-time process analysis and control of machines on a Ford factory floor in Dunton in the UK, investigating the advantages of connecting sensors via a Vodafone 5G Mobile Private Network (MPN) installed in the factory, including low latency, high throughput, enhanced security, and continuous remote connectivity with welding experts at TWI in Cambridge. The network is also integrated with a cloud computing environment where the data is stored, processed and shared with other consortium partners in the project. As Chris White, Manager of 5GEM at Ford, describes in this interview, one misstep or slowdown can impact the entire production line.

What problems are you trying to solve with the real-time process analysis and control project? Although automotive plants are highly standardized, with repetitive processes that are assembling parts made to tight tolerances, we still see variations that can impact the efficiency of the plant and the quality of finished products. These can be as subtle as changes in temperature or even sunlight affecting the cameras used to detect issues on the production line. With this project, we are monitoring the manufacturing of vehicle electrical components through sensors attached to the manufacturing machines. We are also processing the data from those machines in real time to determine if we need to change settings to improve quality. We need to collect large quantities of data quickly and analyze it centrally so that we can look at the end-to-end process – if we make a change to one machine, we have to understand how that will impact other machines in the production line. What kinds of sensors are you using and why? With this project we have 5G-connected sensors monitoring laser welding for hairpin joining and battery-bus-bar joining. Using 5G to connect the sensors wirelessly is ideal because of key 5G characteristics like high data throughput, low latency and the ability to connect with remote experts at TWI in Cambridge, where another 5G Mobile Private Network is also installed. High throughput is important for this project because we need to be able to process large amounts of data quickly; for example, we will check 192 welds per part at least 3 times per cycle, which produces a lot of data. Low latency is very important because, once applied in a full-sized facility, we will be welding hundreds of connections per second – if conditions change, we need to be able to change the settings on the machines within milliseconds to continuously optimize production quality.

Project partner ATS Global is capitalizing on the speed of the 5G network to deliver a hybrid cloud directly to the shopfloor. The platform allows the consortium partners to deploy AI at the network edge, enabling robotic process automation to automatically carry out, repeat and improve processes. Once the product leaves the factory, its lifecycle continues. In the 5GEM project we are using a Dedicated MPN for maximum security, but with other network configurations (see below) 5G‑connected sensors in delivery vehicles can continue to monitor temperature and vibration via the public 5G network to keep track of conditions on the journey. In a sense, the insides of our delivery vehicles are an extension of the 5G-connected production line.

What challenges have you faced so far with this project? Today’s equipment doesn’t come into the factory with 5G-connected sensors pre-installed (although that’s very much the vision for the future). A lot of existing equipment must be retrofitted. Sensor availability is an issue. Whilst IoT sensors are becoming widely available, 5G compatibility will only take off when factories themselves are 5G ready. Factories are incorporating more autonomous equipment that cannot be hard wired, as it will move around. Battery technology and wireless communications will need to advance to keep pace with this trend. In the future, we will need clear standards for equipment manufacturers to use, in order that their equipment is ‘plug and play’ across the whole automotive sector and beyond. We’re using TM Forum’s IoT standards, and TM Forum is leading the standards work in the 5GEM project. What are the key learnings from your project so far? Network installation is ongoing. As ‘non-telecoms’ people, we at Ford had a very limited appreciation of what is involved. The MPN has taken several months to install, partly due to the pandemic, but also because we were ill-prepared for what is required: fibre links, power supplies, GPS antennae locations… It’s certainly not yet an ‘out of the box’ solution, with assembly and testing also taking a significant amount of time. I am sure this will change, as 5G MPNs become mainstream. Installations will need to be reconfigurable quickly. It has been a great learning opportunity for Vodafone and Ford. We have also had a significant debate within Ford about where 5G will fit alongside existing mobile connectivity and emerging alternatives (such as Wi-Fi 6). Ford is studying numerous solutions, and our project aims to inform where 5G will fit. There are many choices to make in how to adopt 5G. Dedicated private networks are suited to this use case, but hybrid or public networks become more interesting when you start to interact with assets that go beyond the factory walls (for example, logistics use cases). We have a non-standalone network, and of course standalone networks are now becoming available with advantages such as lower end-to-end latency – this loops back to the types of sensors and devices that will work in the facility. From a global perspective there are differences in how spectrum is going to be managed, and this will drive complexity for Ford if we want a global solution. What other plans do you have for the 5GEM project? Real-time process analysis and control is just one of several use cases we are exploring. Another important scenario where 5G can help is intelligent maintenance, seeking to eliminate unnecessary downtime of the production line and enable secure remote support by our equipment suppliers using AR and VR. We are also keen to work with the second part of our mobile private network that is based in TWI, home to the country’s best welding experts. We want to use 5G to call on their expertise and possibly even create ‘twins’ of our processes for their experts to work out the optimal welding solutions without stopping the factory itself. To learn more about the 5GEM project or TM Forum’s Collaboration Community, please feel free to reach out to me directly.