At KCOM we talk a lot about the Internet of Things and analytics in the public cloud but they can seem like nebulous technological concepts.  They don’t have to be but context is so important.  The automotive industry offers some great ways of giving IoT some context so please check out my blog about our work with a major car manufacturer on its connected car strategy.

Never underestimate how much inorganic objects can tell you about how they work and how they are used by their owners.  Cars for example can tell you everything you need to know to assess and improve individual components as well as overall service.  Our work with an international automotive manufacturer shows that there are four key areas from which myriad data can be gleaned:

  • Alerts: everything from the washer bottle light to the engine light. All are driven by sensors and all can tell the manufacturer (as well as the driver) vital information
  • Summaries: these give a view of the car’s status as well as, for instance, last journey details and eco scores. Not only that but data regarding the opening of windows, doors and boots etc.
  • GPS: all of the journey data gathered by the car’s GPS system
  • Batteries: this data includes everything around battery performance. We get a message from each car regarding its battery every 30 seconds!

KCOM’s platform handles data from more than two million cars around the world.  That’s around four million messages a minute which is a lot of unorganised info going into our AWS data bucket.  Our platform breaks out the data into individual message buckets as above and that we use AWS Quicksight, an impressive BI tool, to help our customer’s data analyst to be in a position to analyse the data.

While there are undoubtable engineering learnings that can be learnt from this analysis, it’s the driver usage info that is most vital to this project.  Our customer wants to know about driver patterns and how they affect performance.  For instance, how does a particular model perform on the school run in France vs a business commute in Dubai?

Only by understanding user behaviour can recommendations be made for meaningful improvements.  We take queries and build features in Lambda to facilitate this.  For example, understanding that, on average, ten days elapse between the washer bottle light coming on, the bottle emptying and being refilled means that our customer can make changes based on customer behaviour.

So between the sensors and the data analysts there is a world of work to gather, collate and streamline vital information.  Our role is to apply meaning over the top of an enormous raw stream of data so learnings and decisions can be made more efficiently. 

As IoT allows enterprises to understand more about their products after purchase, the opportunities to learn increase exponentially.  Our customer and vehicle analytics project captures data from engines and telematics control unit sensors in connected cars once they leave the forecourt.  When it comes to data, that’s when the fun starts!

If you’re interested in learning more about how the automotive industry is embracing the cloud, check out the keynote at this year’s AWS Summit:

Internet of Things, Cloud