Innotrans 2018 Rail Analytics Survey
Innotrans 2018 Rail Analytics Survey
First off it has to be said what a great event this is for the rail industry, both content and location are fantastic.
So this time round I thought I would give myself a challenge of seeking out all things ‘analytics’. Partly to see where the current ‘digital rail’ initiatives are up to and partly to stop me playing with all the toy train sets and thus not doing any business stuff. Over a 2.5 day period this involved me walking a total of 18 miles, at least that’s what my Samsung health app tells me as corroborated by my aching feet. Even given this vast distance I have to say the survey is not exhaustive in the technical coverage sense but was carried out in more of an impulse shopping type manner.
Check #railanalyticshunt on Twitter for snapshots of the 2.5 days.
Setting the scene for the survey:
There has recently been an increase in the use of the term ‘analytics’, according to my other non-exhaustive survey of LinkedIn posts. This, along other words such as ‘digitalization’ and ‘machine learning’ and ‘algorithms’ got me wondering what is going on in the rail industry. Has is uncovered a new mathematical/computer science dictionary and suddenly found a new set of words to flummox us with. So I thought I had best try and put some science into the game and do a survey of what is actually going on beneath the words.
My chosen approach was to focus on ‘analytics’. This was for a couple of reasons. The first was to take the focus off all of the shiny widgets on display and the second was to try and get to the heart of what the calculational landscape was looking like.
To this end I defined 3 rough classes for what constituted ‘analytics’, these are;
- Class A – just analytics, business as usual, single or ‘few’ data stream analyses,
- Class AA – advanced analytics, more than business as usual, with analysis involving multiple data streams,
- Class AAA – absolutely advanced analytics, nirvana, with complex analysis on multiple diverse data streams.
All of which have value of course, the classification wasn’t like high school grading but more along the lines of the way batteries are marked.
The survey findings:
I hit the wall pretty early on. It seemed like every stand had the words ‘intelligent’ and ‘innovation’ somewhere on their banners. So what I thought was going to be a simple wander through the stands picking out likely candidates assessment was going to involve a bit more brain engagement. Fortunately the first stand I decided to have a proper look at kicked me into life. That was the Siemens stand. Very impressive offerings which have a deeper darker side to them. Siemens, I believe, are well on the way to becoming the Google of rail data. They are building a data ecosystem with associated access protocols to facilitate app development. They themselves have built passenger journey apps which are delivered to your phone and are working on enhancing these to provide better planning around potential train disruptions. The dark side being their control over access to this data. I classed this as an AA on my rating scheme for use as a basis to compare others too.
The vast majority the other candidates tended to focus on BAU, class A, analysis. Essentially time series type analysis on condition data looking for trends and/or changes to trends.Lots of good stuff on shown particularly around ‘predictive analytics’ or reliability centered maintenance in old currency. I would say 85% of those surveyed were in this category. A secondary factor in this class was the ‘sexyness’ of any associated hardware. Probably the sexiest bit of kit was the KONUX (slightly unfortunate name sounding like con-us) a shiny yellow thing, a bit Apple-esque in nature. They wouldn’t tell me what was in the box but it was essentially a range of motion/vibration sensors. The whole assembly is then bolted to the track at strategic locations to look for changes in motion profiles.
The class A category also had a number of vision based analytics entries in it. Very sophisticated tracking of object on CCTV images. Lots of companies providing similar offerings. There was also a range of VR products for helping improve maintenance training/delivery. Personally I still don’t quite get the use of VR. Seems very gimmicky still. My colleague tried on a Hololens variant and just ended up making himself sick in the process. I’m sure it will all come good in the end.
I then stumbled across Amazon, well not quite, but a close resemblance called ‘stationOne’ which is an Alstom offering. Which is, quote, ‘a global digital platform to sell and buy mobility products’. Which sounds a bit like its selling scooters but in actual fact it’s a site to buy all of the widgets needed in the mobility (rail) arena. Essentially Amazon for rail techies, they are probably using AWS to delivery (sorry). I am assuming they are also using the back end analytics associated with this type of offering. If this is true then this would be a class AAA candidate.
In summary, survey details shows
- 85% of analytics stands are class A
- 10% are class AA
- 5% are class AAA
Clearly plenty of room at the top end of analytics for those who want to go there. One issue that will raise its head soon is the access to data sources. Most are tending to lock down their data. Oh sorry, you did say it was all open source, I remember that, but only open to those who pay the entry fee. It’s the business case stupid. Ah, now I understand.
The big German and French players dominated the analytics space at Innotrans 2018 showing off class AA and AAA capabilities. There were some outfits honing out niches in the analytics space mainly in class A. There’s clearly vast quantities of money being pumped into rail digitalization (or digitalisation), which seems to me to be just digitisation but sounding a bit cooler.
I’d found Google, Apple and Amazon during the survey but where was the rail version of Facebook I was wondering while munching on a currywurst. Then I realised that space was actually taken up by Innotrans itself. Which has grown tremendously over the past 10 years with no sight in end, trendy meeting places, networking with old colleagues meeting new people. Facebook-like indeed!
I have thus revealed the 4 Horsemen of Innotrans 2018.
Next update 2020!
Dr Gary Bamford CEng, FBCS, CITP
Advisory and Risk Management Solutions Ltd
@grandwizz #railanalyticshunt for more info.