Leveraging on Data
The Digital Transformation accelerates the generation of data and companies are learning to capitalise on those data, sometimes clashing with fuzzy regulation and unclear definition of ownership.
Take the example of Tesla. Tesla have been designed through computer modelling and this process generates a digital model of the car that is used for simulation and for interaction with component suppliers. The model is then turned into software for the manufacturing of the car, to create parts and assemble them into the finished product. All the manufacturing phases are recorded and are added to the model of the car. At this point there are many instances of that model, each one mirroring a specific car (there might be difference, in colour, optional features, even in components that might be provided by different suppliers and of course in the robots and workers that manufactured a specific car. Once the car is sold sensors embedded in the car keep track of its operation, where it goes (Tesla is adding data on 1 million miles of road driven every 10 hours!, on November 2018 they reached the thresholds of 1 billion miles recorded), when it is recharged, how it is used… All this information is harvested by Tesla for proactive maintenance (to detect possible malfunction) as well as to detect issues in the manufacturing that might become apparent only at a later stage. As an example, a few cars after some thousands miles on the road showed an abnormal vibration detected by the sensors on the left suspension.
Through data analytics AI software was able to pinpoint the problem in an asymmetry in the fastening of bolts on the two sides of the car assembly line, with one robot placing a higher pressure than the other. At that point the software inserted a notice in those cars digital twin (the car data sheet) so that once one of them was taken to a repair shop the notice will be read and recalibration executed. All of this without the owner ever becoming aware of the problem (the vibration was so tiny that went undetected by the driver).
Also, all data harvested from driving, including the data generated by the digital cameras used by the automatic driving assistant (autopilot) are continuously analysed to improve the autopilot using machine learning. In other words all Tesla cars are getting better every single day by leveraging on the experience of each and all of them.
Data are so fundamental for Tesla and its business that the company has been defined “The Data Company”.
Although operating in a tiny vertical sector, the one of private transportation, Tesla is gathering a massive amount of data that can generate plenty of metadata, like:
- Where are people driving, when they are driving, how they drive, with whom they drive (this information is easily obtained by sensing the smartphone on the car, via Bluetooth). This information places Tesla in an ideal position to evaluate risks and therefore to offer competitive car insurance;
- What are the road conditions and how these condition change over time, whether there are potholes requiring maintenance and so on;
- What traffic is on a particular road at a particular time, detecting traffic anomalies versus the usual pattern. Get information on other cars driving around (their number plate is easily captured by the Tesla camera and it would be straightforward to map, given sufficient time, the whereabouts of many other people in a given area);
- Who is walking around and who is getting in and out of shops and homes. A bit of data correlation would make it easy to find out about other people habits in a specific area;
- The car suspension can provide data on the “weight on the car” meaning that over time it is possible to gauge various parameters about the driver, the passengers and even the groceries being transported;
Notice that what listed above is not stating that Tesla is actually gathering these data, just that it would not be a technical problem to harvest them. As a matter of fact Elon Musk has stated several times the advantage given to Tesla by the use of Big Data and AI Data Analytics and once even voiced the idea of using these data for selling services. That resulted in a backlash by several Tesla owners concerned of the violation of privacy on the one hand and on the exploitation of their data to make money on the other.
Clearly Tesla is the perfect example of a company born out of the Digital Transformation of a classic business. It is not just the shift to electrical power engines, it is the use of data that is making Tesla worth more than Exxon in the stock market. Its turnaround is way, way, lower than Exxon, yet the stock market expectation for a bright future is all on Tesla side.
Tesla is just an example of the power of leveraging data for a manufacturer:
- fine tuning of manufacturing based on operation feedback
- continuous improvement of products through data analytics from the all products sold looking at the way they are used
- development of services flanking the product
- offer new services in other market segments that are in principle completely outside of the company business (like providing info to municipalities on proactive road maintenance).
Tesla has been used as an example. According to a new McKinsey report the monetisation of big data harvested from cars can lead to a 750 billion $ market in 2030.
Today there are many companies that are leveraging on the result of the digital transformation of their business and exploit data to increase revenues, decrease cost and strengthen their market position.