5G & the Factory of the Future

You may have noticed recently that operators and vendors have started to talk up "Factory 4.0" as a banner heading for another use for 5G.

This envisages using 5G as the bedrock for updating factory facilities across the globe, to support robotics, automation, artificial intelligence and more. In fact, when I say "bedrock," that even means using 5G wireless connectivity to get rid of the wiring in such facilities. (See Nokia Reveals Future X Network Project and Ericsson's CTO Talks Up 5G Opportunities.)

Factory 4.0 applications appear particularly suited to high-band millimeter wave 5G networks that can deliver massive data (1-Gbit/s and up) flows over a short range.

Want to know more about 5G? Check out our dedicated 5G content channel here on
Light Reading.

One key aspect of the 5G spec, however, that needs to be completed before many Factory 4.0 applications can be supported, is called Ultra-Reliable Low Latency Communication (URLLC), or guaranteed network latencies of a few milliseconds or less. This is due in the 3GPP Release 16 -- a.k.a. 5G NR Phase 2 -- due to be completed December 2019. This likely means the updated standard will get into commercial usage sometime in 2020 at the earliest.

URLLC will be necessary for applications such as precision robotics and automated delivery. It will also be necessary when people want to use thin client devices, which rely on millisecond connections to the cloud so that "dumb" devices can handle computing tasks. (See Could 5G Revitalize the Thin Client Market?)

Naturally, this also applies to the push towards automated vehicles.

So, it seems that the factory of the future could become much more of a money-maker for carriers and wireless vendors, but not quite yet!

— Dan Jones, Mobile Editor, Light Reading

Page 1 / 2   >   >>
DanJones 11/16/2018 | 4:19:52 PM
Re: Wishful thinking? Nokia, was using a couple of Bosch robot arms for demos last week. They didn't want photos taken,which was wierd.
Duh! 11/16/2018 | 2:55:43 PM
Re: Wishful thinking? I'm skeptical of this solution, not the use case.  The latter is clearly being driven by manufacturers and industrial machine makers, not telecoms. Robert Bosch is apparently leading the charge. Quick rebuttals to your points:

1)  Apparently, 20 ms communications latency is sufficent for many of the applications they have in mind.

2) I don't get your point. Embedding comms modules is not a big deal. Designers and/or standards folks certainly have a choice of pluggable form factors, if that's what's needed. Software defined radio is another possibility if there are multiple standards.

3) This is a global movement. As it happens, manufacturing is coming back to the US and Europe, sans traditional manufacturing jobs. This is likely a part of that.

4) The presentations I've heard talk in terms of months or years, but with minimal disruption during a change-over. Automobiles has been given as an example. Another example would be moving capacity around a high-volume/low mix factory.

On your main point, I agree. The industry tends to create solutions looking probems.  Sometimes that works out, sometimes not. Industry execs are also inclined to a certain solipsism, insofar as their desire to get inside the demarc and try to compete with domain-expert companies with end-to-end solutions.

Let's take this offline, if you want to continue.


brooks7 11/15/2018 | 1:15:12 PM
Re: Wishful thinking? Gabriel,

Let me give you a better notion of my skepticism.

1 - Latency:  I worked for a brief time in Real-Time Controls a long time ago.  At that time, we were working with 1 msec full loop control cycles not 10 msec transmissions.  The industry has changed, but I am guessing the control cycles have not gotten longer.  That implies smarts closer to the sensors - not farther away.  So centralized orchestration at some level is needed.  But does a $250 1RU high server in a box really cost so much or create so much burden?

2 - Availability:  I repeat that the dongle thing (otherwise the machines have to have native 5G connectivity) is not going to work in any kind of volume.  

3 - Greenfield Only:  Only brand new factories (which one assumes are all in China right?) need apply.  That greatly restricts the market and one wonders if you were building a $500M facility that you would bet it on a new, unproven model.

4 - Reconfigurability:  Every day?  Many times a day?  Sounds like a low-volume, high mix facility.  Maybe there is some advantage in moving machines, but it sounds like a lot of labor instead of just hiring some guys to move pallets.

What I am trying to say is that this seems like yet another idea to take a perfectly good technology and PUSH it onto a market that does not seem like a great fit.  One of my complaints about the way this site tends to work is that we are our own echo chamber.  Carriers propose ideas to vendors.  Vendors make equipment to meet those Carrier needs.  Nobody talks to actual customers and see how they solve the problem today and that if this solution actually improves their business.  Suppose 5G (or WiFi whatever) makes this 1% better?  Then it will never sell.  It will prototype and people will spend money on it.  I think it is the job of vendors (in particular) to be skeptics about any new opportunity like this, until they have end user conversations that make business sense.


PS - Hell we did this at AFC with our DSL Everywhere initiative in 1999.  Our biggest issue is that it was just a PR effort and then people wanted to join the group, set standards, and get products into the field.  We had to backpeddle like crazy.

Gabriel Brown 11/15/2018 | 11:06:52 AM
Re: Wishful thinking? Ha! Oh well...
brooks7 11/15/2018 | 11:01:37 AM
Re: Wishful thinking? Actually that site makes me believe that it is just a marketing guy (okay a bunch of marketing guys).  5G marketing guys.

Gabriel Brown 11/15/2018 | 4:02:14 AM
Re: Wishful thinking? It's good to question and a degree of scepticism is warranted, but this is more than something "made-up by a marketing guy".

The 5G Alliance for Connected Industries and Automation (5G-ACIA) is a good start https://www.5g-acia.org/
brooks7 11/15/2018 | 1:56:11 AM
Re: Wishful thinking? The assembly lines that are massive conveyer systems bolted down and the robots that are the size of cars?

I get the compute part, but if we are centralizing it....uh why are we moving it?

I guess I am saying this sounds like something made up by a Marketing Guy who has never actually worked in a factory.


Duh! 11/14/2018 | 5:15:04 PM
Re: Wishful thinking? Probably not heavy machinery. More like robots and assembly lines.

DanJones 11/14/2018 | 3:24:40 PM
Re: Wishful thinking? Well, reconfigurably can also include moving compute power (edge, cloud etc,) to handle specific tasks.
brooks7 11/14/2018 | 2:04:22 PM
Re: Wishful thinking? Rapid Reconfigurabilty using machines that way several tons....okay sure.

So, its a brand new factory and all the machines have 5G built in.  Because if they have dongles it is a big latency problem right?

Seems like a long way off.


Page 1 / 2   >   >>
Sign In