What Might the Demand Be for 5G in Manufacturing?
Many of those working in and around mobile telecommunications are increasingly of the view that one of the key applications and opportunities for 5G will be the so-called Industrial Internet of Things (I-IoT).
The view is that industries such as manufacturing and logistics would find the new features of 5G, especially its promised low latency, particularly useful. There have been some case studies which suggest that productivity in some factories might be improved if certain machines could have wireless connectivity to a control center that could react very quickly to signals from the machine, perhaps indicating a potential failure or error condition. These have included Bosch in the UK (where they make home boilers) and car manufacturers in Germany.
It seems intuitively obvious that:
1. Some machines would work better if connected to a central control and monitoring system.
2. If this connectivity is wireless, it allows the machine to be moved and avoids the difficulties of laying wires in a manufacturing environment (but noting most machines will at least need wired power and not all of them need to move).
3. Lower latency, higher data capacity and better reliability are all helpful.
Equally, in many cases, wires will be acceptable. Alternatively, simpler wireless solutions such as WiFi, self-provided IoT (such as LoRa or Weightless, a development with which I am closely connected), or proprietary solutions will be able to provide the capabilities needed.
If this 5G opportunity does exist, then how large is it? Interestingly, given all of those who have proclaimed that I-IoT could be the making of 5G and the source of the much-needed new revenue for operators, I have not seen any publicly available estimates for the size of this market. What follows is meant as a starting point for debate.
According to the statistical office of the European Union (EU) -- see this May 2019 update -- there are 16,100 manufacturing enterprises across Europe that are classified as large (in that they have more than 250 employees). My assumption is that, for the most part, it will be the large organisations that will be interested in 5G, since 5G solutions will likely be expensive and relatively complex compared with WiFi. In the EU, these companies tend to be focused on the manufacture of coke and refined petroleum products, motor vehicles, tobacco products and other transport equipment.
With 28 countries in the EU (for now at least…), the average country has 570 large manufacturers. Of course, these will be very unevenly distributed, with Germany likely have much more than this. And outside of the EU, it seems likely that there are many more manufacturers in the Asia-Pacific region.
Let us make some simple assumptions. Firstly, that 50% of these will benefit from having wireless connectivity (the rest either being content with wired connectivity or not needing any significant connectivity). That leaves 285 per country. Most will find WiFi, or similar, perfectly adequate, so let us assume that about 20% want 5G connectivity, primarily to achieve latency targets. That leads to potential demand from 57 per country. Most countries have three or four mobile network operators (MNOs), so that averages out at around 15 manufacturers per MNO.
How much could an MNO charge? If it charged a lot more than the cost of self-provision, then many manufacturers will decide 5G is not really needed. A self-provided WiFi system might cost about £50,000 (US$61,200), amortised over 5 years. So £10,000 ($12,240) per year might be reasonable on average (of course, there will be some that will pay much more -- perhaps some of those German automotive manufacturers). Overall this equates to a potential revenue stream of about £150,000 ($183,500) year per MNO. That amounts to less than 0.01% of their average current revenue. Even if I am out by two orders of magnitude, that still does not change MNO revenues materially.
Also worth considering is that if the market is this small, then the major suppliers may not produce the type of equipment that would be best suited to this deployment, leaving MNOs to use standard 5G solutions that may be expensive or over-engineered.
An alternative option is that these industrial companies deploy their own solution. That would be closer to what typically happens today, and may be preferred by the companies that can then have full control over what will become a critical part of their process.
However, that is complicated by a lack of access to radio spectrum. In most countries there is no 5G spectrum available for non-MNOs: Germany is an exception, with the proposal to set aside some spectrum for manufacturing, and some countries are considering shared access to 5G spectrum, but few have implemented it. Without spectrum, self-deployment is not possible. If the spectrum position is resolved, then this approach might generate some revenue for the equipment manufacturers, but none for the MNOs.
So is 5G I-IoT really so important? It might well be valuable for a few manufacturers or logistics companies, but it does not feel like it will change much for the 5G community, except perhaps in a few countries with a larger number of leading-edge manufacturing companies.
I-IoT is likely to be important, but may generally use existing wired and wireless technologies and not require much support from the telecoms industry.
— Professor William Webb, Consultant and CEO of the Weightless SIG