Tesla, big data and industrial disruption

Author

Paul Gittins

September 18, 2015

I have a 10-year old Audi sat on my drive – it’s been a good car and has had one recall and 2 two minor software updates to the sat nav in 10 years. Neither brought me any more functionality.
 
Someone recently sent me the Tesla software change log…  pages and pages of improvements, with a change log showing functionality updates every few days…
 
… which lead me to realize that Tesla is showing all the signs of being a fundamental disruptor to industrial manufacturing – now complemented by the recent announcements into the world of domestic and grid batteries.
 
Let me explain my rationale.

Implications to data

It is likely that Tesla has taken a “digital first” approach to the whole customer lifecycle and modeled customer behavior on all streams of social media as well as their configuration process. It’s likely this extends.
 
In this new world (and Tesla is not unique in telematics) – car manufacturers collate data via an opt-out “Vehicle Telematics Subscription” model.  They are using telematics to batch stream key data points (not a full data dump from the car with every data point as I understand it) via 3G & wifi[1] to some form of data lake at Tesla.  Interestingly, they replaced their SAP ERP with this model, a view which is entirely in line with our thinking with the Business Data Lake* that we launched in late 2013. From this lake they are creating distilled data and analytic views, focusing on long term analysis  - not just to create a radically improved closed loop engineering cycle, but also to provide a continuously improving customer experience.

Implications to manufacturing platform engineering

This implies a couple of really key shifts in car development vs a traditional manufacturer:

  • A software development cycle that is very tightly integrated with the “test mules” & platform development cycle so that improvements can be quickly reflected in the production cycle
  • Near-real time customer usage information that changes a production change cycle from months/years to weeks
  • The marrying of very rapid development cycles with safety critical development… normally a dichotomy?

Implications to Tesla product development

Beneath the surface, Tesla seems to have created: 

  • A digital-centric business model enabling a continuous improvement program of the customer experience, device features and underlying hardware control layers, linked to a level of vertical integration in core technologies
  • Products that from day one and through thousands of users’ data, provide them with an ability to develop and roll out incremental car and customer experience features at pace, impacting every aspect of the customer lifecycle
  • Owners can request new functions and potentially gain them in weeks, not years, matched with the ability for the manufacturer to extend their test capability by creating a community of early adopters for new features rewarded with continued early access

The upstream implications are as significant as the customer experience. 

  • True predictive maintenance fully integrated into the customer service experience. Manage customer issues before they impact satisfaction.
  • Cars have many, many components and Tesla’s work is integrated at the manufacturing level but not for all car components. As they expand the data sets that they can capture, they will end up with an ability to provide a fully federated insight supply chains:
  • The ability for the car (or an autonomous agent on request of the car) to sense an abnormality and request an analysis on that abnormality to produce not just an alert – but also a “next best action” for the supply chain and the customer – e.g. automated increase of that component, order ahead of failure, automated booking for repair
  • The ability for component manufacturers to request anonymized data insight into their components – enabling improvements whilst providing assurance over customer privacy and security but enriching their lifecycle knowledge of that component.

 The key for the consumer is this:
 
“Your next car will be even better when you come to sell it”.
 
This is the heart of the change.

The used car that you eventually sell will be more efficient, go faster, and be more feature rich than the one you bought, and it was the customer sharing data, together with big data and analytics that has helped make it so.
 

Tesla as a new industrial model

Tesla may become wider disruptor of any power intensive activity – cars, new homes, small scale industrial, etc with the value being in both the vertical integration (are batteries the end game, or the start point?), the ability to marry multiple data sets to drive a transformative experience throughout the customer lifecycle.
 
Tesla is potentially the most disruptive digital “large device” manufacturer yet to emerge, aside perhaps from GE with their Predix platform (though arguably their challenge is retrofit and incremental services development).
 
Watch Tesla… The future of automotive is very, very exciting.
 

* For clarity, there is no implication that Capgemini has performed any work for Tesla; it is interesting to see that a form of parallel evolution is at play.
** Declaration of interest; I don’t own one, but if Tesla is reading, I’d happily try one.


[1] Ubiquitous high bandwidth via telematics will no doubt remain an issue for some time to come, but the ability to upload a batch via wifi is in place.

This article was written by Paul Gittins from CapGemini: Insights & Data Blog and was legally licensed through the NewsCred publisher network.


Comment this article

Great ! Thanks for your subscription !

You will soon receive the first Content Loop Newsletter