Why rail operators don’t need 4G for remote monitoring— or at least not in the way they think they do

Charles-Henri Mousset
Railnova Blog
Published in
6 min readApr 28, 2017

There has been lots of buzz around 4G in the recent years and how to apply it for rail passengers WIFI, railway IoT and remote condition monitoring. Everyone wants remote monitoring systems to be future proof, and no one would buy a 3G iPhone today, so why would you still buy a 3G remote condition monitoring device instead of a 4G one?

In my position as hardware product manager at Railnova, I need to decide whether we build our next generation remote condition monitoring device for 3G or 4G (or both), taking into account the manufacturing cost, coverage and SIM operating costs for our clients. I’m not a telecom expert, but I’ll share my views on the topic here, since I’ve had to answer this question at least 5 times (!) recently.

Let’s dig a little deeper in the quite opaque “telecom” world and try to understand what’s the deal:

1) 3G won’t disappear

First of all, 3G won’t disappear. Network operators actually continue to add 4G/3G retro-compatible (and also 3G-only) Base-Station Subsystems for a simple reason: 3G is currently still cheaper than 4G for rural coverage.

Not only are 3G Base-Station Subsystems (BSS) cheaper, but they also have better coverage (if the network speed goes up, the range goes down).
To provide the same coverage at high speed you’d need a lot of extra BSS. Now, if you’d add the 4G markup to those Base-Station Subsystems, your rural 4G coverage infrastructure would cost at least twice as much as the 3G one, while revenue might not double, making it a difficult case for telecom investment. Also, 3G still looks to be the the preferred replacement for 2G Voice call worldwide, so network operators have an additional incentive to keep investing in 3G.

In my opinion, 3G won’t disappear in the next decade, and will probably survive way over 2030.

2) 3G and 4G offer a different coverage

In Europe 4G and 5G -faster wireless networks- are mainly developed to solve the problem of congestion in cities. Keep in mind that it’s a radio network: everyone on the network shares the same frequencies. Technically, this means there is a hard limit to the total bandwidth thin air can see passing through.

In big cities, with tons of people streaming full HD and sometimes even 4k videos on their cellphones, this hard limit can easily be blown with 3G. So, an urban mobile Game-of-Thrones-watcher should probably buy a 4G phone (and leave the 2G & 3G radio frequencies free for the farmers and trains that need maximum coverage).

This is reflected on France’s current 4G coverage, which is much less than 3G in rural areas:

4G coverage in France, February 2017, source: opensignal.com
3G coverage in France, February 2017, source: opensignal.com

3) Can 4G be of any good for rail operators in the context of remote monitoring?

Considering that a) the average remote condition monitoring system creates 100 to 300MB per asset per month and b) 3G speed can support ~36GB per month (based on 20% practical speed of a 5Mbps connection, 12h of asset utilisation per day, 30 days/month), it’s hard to justify the need for 4G for train operators at this moment.

3G already offers 10 to 100x more bandwidth than currently needed by most monitoring use-cases: even if we wanted to upload the full real time content of the MVB bus of a train (we would reach ~139GB of monthly data: 100Kb/s on the bus * 3600 * 12 * 30 / 1000000), we could easily reduce the 139GB of raw MVB bus data to 15GB using a simple compression strategy, which would then also be supported by 3G.

4G on the other hand, requires a different GSM module and a more complex antennae arrangement (‘RX diversity’) which doubles the antenna costs and placement constraints, all adding up to 15% extra cost on the device manufacturing costs.

So, short answer to the question above: No, 4G-only subscriptions are currently not an interesting option, as telecoms won’t increase their rural coverage soon and 3G offers a good, future proof solution for remote monitoring projects.

Long answer: Technically, since 4G provides much more bandwidth than 3G, the cost per MB is much lower for network operators. In the near future, it’s possible that 4G-only subscriptions could cut prices by a 2-factor. However, for this to work, the 4G coverage would need to be continuous when the train is moving through rural areas. As that’s not yet the case, another solution is needed.

4) Multi-SIM to the rescue

Let’s imagine you need to transmit tons of data (for example: asset data, CCTV downloads, raw train bus information etc.) from your assets to your servers or big data systems.

3G and 2G would be perfect to transmit real-time data about your asset condition as the coverage is good in both urban and rural areas. 4G can be used for larger data sets, but would be of no practical use when the asset is moving across rural areas (no coverage). However, when the railway asset is parked close to urban areas it has access to 4G and can use it to transmit plenty of GB.

You then have the best of both worlds: superb 3G/2G coverage for always-on tracking; and cheap data to transfer the bulk you don’t need ASAP with 4G.

This differentiated approach can be compared with the good old day of ‘GPRS + WIFI solutions’ in the early 2000’s: GPRS was used to transmit a small amount of data while on the move, and WIFI was used for full download when the train came back into the workshop in the evening.

To our knowledge, no network operator offers a service with both top-notch coverage and cheap 4G rates yet (but 4G-only sim cards plans are now showing up). Unfortunately, that means you are stuck with 2 SIMs and 2 contracts per train, but, on the bright side, you get maximum flexibility in data volumes and coverage.

Also, a 2-SIM setup gives you bargaining power over network operators: usually, swapping SIM cards induces big switching costs (SIM cards, managing the swap campaign, asset downtime, contract overlap, etc.), but with the 2-SIM setup, switching from SIM A to SIM B is possible and just a matter of remotely updating a configuration. This gives you some degree of commercial flexibility towards telecom operators.

In conclusion

4G speed for railway IoT and remote monitoring data transmission might be interesting in urban areas, but only if you have a real need for big data volumes (like CCTV). If you operate in both urban and rural areas however, you’ll need a 3G/2G failover for continuous monitoring and the best coverage at any time.

A proper multi-SIM system will help you take advantage of 2 specialised GSM subscriptions (4G being high volume with low coverage, and 3G being lower volume with higher coverage), and offer you full coverage anywhere and anytime. That’s why I decided to implement a dual SIM module in our next generation remote monitoring device, where we can use each SIM in 3G and 4G.

Want to learn more about how to get the best out of a multi-SIM system? Be sure to get in touch with us or meet us at Transport Logistics in Munich from May 9th to May 11th, we’d be happy to share our experiences with you!

Like what you’re reading? Here’s more good stuff:

--

--