Seamus K - Irish tech industry expat living in Sweden.

Month: February 2020

A cautionary tale about IOT. Sorry, about poor design decisions.

This story turned up in my RSS* feeds this morning.

This has to be a nightmare story for someone working in PR. Your product is being used by a journalist. It fails in a major (if not life threatening) way. Your customer service people badly mismanage their call for help. The story goes viral.

PR person – your assignment is to suggest strategies for salvaging something from this shit storm.

Now that the story has grown legs, there will be plenty of people whining about 5G, IOT and the like. Except, I would say that this is a problem for the car company not with mobile connectivity. Because as someone who works in the mobile technology business** I can tell you that this is not an unknown issue. I was working on it over 10 years ago. 

Back then I was involved in a project around connected cars involving my employer and a luxury car manufacturer. This company had an emergency assist system in their high end autos – if the car conks out, or just runs out of petrol, then hit the button and you will be connected to help (for a very reasonable monthly fee). They wanted to make sure it was as robust as possible, and were interested in seeing what other services could be built on top of this once the car was connected.

There are a load of interesting challenges you have to connecting a vehicle (I worked on trucks as well at the time):

  • Vehicles tend to have long lives compared to mobile devices. So any technical solution had to be able to last for at least 10 years without maintenance, without going obsolete (is that frequency going to stay allocated to that tech for a few decades)?
  • It had to survive being stuck in a box somewhere that might be subject to extremes of heat, cold and vibration like say the engine compartment.
  • High end cars often have filament heaters in all the windows which effectively turn the car into a faraday cage. We could use an external antennae for the car system, but passengers cell phones will probably take a huge hit to their signal. 
  • Emergency systems tend be needed quite often in remote places where coverage is more of a challenge. And they were worried that someone paying a monthly subscription for this service would have no signal, while their au-pair’s pre-pay phone would. Embarrassing.

It became clear that while it was the dawn of the 3G age, GSM was the way to go for the moment. But it also was clear that any service needed roaming – both international roaming (big problem in Europe where you can get through 5 countries in a day) and also national roaming. And there you step from a technical problem to a business and economic one. Which of course is why I was involved…

At the end of the day perfect coverage is never going to be possible (ever tried to get a cell signal at the bottom of a remote forested valley, or the basement level of a car park, or even just in many countries metro systems)? Any system needs to be able to handle this. I think there is a technical term for this – oh yeah “Failsafe“.

So while my employer, and I, and a very large global industry base, will continue to push IOT (and 5G and ubiquitous connectivity) forward, we understand that it has its limitations, like any technology. You need to be ready for when (not if) it reaches its limits. And in the case of the connected rental car this was a predictable failure. The system should have been better able to handle a loss of signal. The article says the company has some approaches for this, but they were not robust enough to keep this journalist moving.

And as for the support person who suggested to a journalist for a globally read publication that they should spend the night sleeping in the car in the woods

Oh dear. Oh dear oh dear…

* Do you remember RSS? I do, and still use it all the time. Because I am one of those weirdos who think I should decide what new stories I want read. I remain unwilling to let Google or Facebook or Twitter decide that for me. Give me the firehose of 100 and I will pick the ones I want dammit!

** My employer is one of the big three mobile technology vendors. Which means a) I have an interest c) actual knowledge about this topic c) strong opinions on it 🙂

[2020 Target: 4/52. 1640/25,000]

Facebooktwitterredditpinterestlinkedintumblr

© 2020 www.sliabh.net

Theme by Anders NorénUp ↑