Building the agile factory of the future
I’ve been seeing images of gleaming yellow and orange articulated robots jabbing at half-assembled products for most of my life — decades, anyway. The idea of automated assembly isn’t something new. What is changing, however, is the extent of automation in today’s factories, and the way digital technologies are making robots smarter, more flexible and more productive. A crucial part of this story is also the way in which mobile wireless technologies are making an entirely different, more agile factory possible.
The digital transformation of today’s manufacturing, which is sometimes called Industry 4.0, is the result of several innovation trends coming together: low-powered Industrial IoT (IIoT) sensing, AI and machine learning, edge cloud computing and new industrial-strength private wireless communications. In this post, I’m going to focus on the evolution of wireless technology, but I will also touch briefly on the others.
Digital automation has been used by manufacturers for some time, but the basic set up of assembly line process has remained a fixed configuration. For example, robots need to be placed precisely. Robotic arms are limited in terms of their range of motion, and it is best to have them move mostly in one direction. Designing for automated assembly requires a lot of time ensuring that the assembly line orientations are optimal. This can result in tremendous productivity, but agile, it is not. Moving these assets from one production line process to another is costly. Re-cabling is one of the biggest expenses with estimates ranging between €500–1,000 per connection.
Now, imagine if the robots could be moved flexibly around the production floor, orienting themselves to the piece being assembled based on sensor input. This would mean tremendous agility and the ability to re-tool factories in very little time. Today’s robots can be designed to be very conservative in their use of energy so that they can be self-powered. They can be armed with sensors and cameras that, like autonomous vehicles, can respond to workers or other machines coming into their operating space.
To support this kind of mobile work environment, the compute resources required to support the software analytics that drive the robots need to be on the edge, what is called multi-access edge computing (MEC). The local processing on the edge is used for handling raw data, which is the bulk of the real-time monitoring, AI/ML and analytics work — a centralized cloud adds too much latency and wide area connectivity costs. Machine reactivity to inputs also requires the low latency of edge processing. When a pattern is established that can be used to train the robots for improved performance, that information can then be kept in the cloud.
This means that there is a lot of critical operational data flying around in the factory that requires high reliability and security, supports low latencies and mobility. Until recently, that posed a problem for early generation wireless technologies, such as Wi-Fi and Wi-Max. They are not designed as mission-critical communication technologies, and they don’t support reliable mobility: the hand-off between Wi-Fi access points is not always seamless. Their coverage is spotty and, although they are capable of low latency, high bandwidth performance, it is not predictable or reliable, especially if the number of devices suddenly increases in a coverage area. And, finally, they are notoriously easy to hack.
Industrial grade 4.9G and 5G private wireless networks, fortunately, solves these problems. Based on standards-based, field-proven mobile technology, these industrial wireless solutions tick all the boxes for the agile factory. They are designed to provide coverage to high-speed trains, much faster than any robot will move. They provide very high bandwidth (10Gb/s+). Private wireless 4.9G networks can provide latencies in the 10–20ms range, and 5G, as low as 1–2ms consistently and reliably.
4.9G which is an enhancement over basic LTE is very secure. These technologies were first developed for public mobile networks and, because of the standards established for almost a century in the telephone system, they were expected to be highly reliable. After a decade of use in highly exposed public networks, there are no reported examples of a commercial 4G/LTE networks being compromised.
The radio spectrum that 4G/LTE has typically used has been monopolized by mobile operators until recently. Governments are now releasing shared and dedicated spectrum for industrial use. Thus, although 4G is a decade old, it is only recently a practical and affordable technology for industrial use. When the standards are completed for industrial 5G in two to three years, it will offer many specific technologies for industrial use, such as the ultra-low latencies noted above and access to an even wider range of radio spectrum.
The availability of this next-generation, industrial wireless will create a lot of new use cases for advanced manufacturing. Along with flexible production lines, it will empower workers with AR and VR technologies, keep them safe with connected PPE, and allow for intelligent AMRs that can use context-aware open-path navigation to route around obstacles, workers and other robots.
Ubiquitous, highly reliable wireless connectivity is the key technology for connecting and digitally transforming the entire factory. It will enable digital twin modelling of the end-to-end workflow to better understand the efficiency of machines, workers and entire production lines. It can also provide data that can be used for preventive maintenance and quality control.
The level of disruption that Industry 4.0 technologies will create is unprecedented. I’m personally excited to see where this will all go. Manufacturers that do not move quickly to digitally transform and automate their operations will find it increasingly difficult to compete with more agile competitors. There are still hurdles to overcome in our evolution to this fourth industrial revolution, but key pieces are in place, such as AI, machine learning, edge-cloud computing and industrial IoT. Industrial-strength wireless — 4.9G today and 5G tomorrow — is the connective glue that will bring it all together.
To learn more about the coming together of AI, machine learning, Industrial IoT and industrial wireless communications based on 4.9G and 5G technologies, read the new eBook: Are you ready to build the digital factory of the future?
Share your thoughts on this topic by joining the Twitter discussion with @nokiaindustries using #5G #industry40 #IoT #PrivateWireless #DigitalFactories #DigitalFactory #manufacturing