The future of the connected car and autonomous vehicles lies in the effective management of data. In an automotive world where cars are tapped into the internet and part of a co-ordinated infrastructure of billions of devices, getting the right data to the right place at the right time, will be critical to delivering the connected car services; whether that be integrated safety, collision avoidance, autonomous driving, fuel and route efficiency, predictive vehicle maintenance or cutting edge infotainment, says Michael Ger, general manager, industrial manufacturing and automotive solutions at Hortonworks.
Powering with data
The power of data in the connected car, broadly, is two-fold; firstly, data allows connected devices within the car to respond automatically and provide instant information to the driver to improve their experience, secondly, the data collected from the devices within cars can be fed back to manufacturers for analysis and improvements, potentially, to create the autonomous vehicle of the future. These two use cases must work in parallel to be effective in the long term, which can prove difficult, given the extent of data already being generated by connected cars.
With up to 25 Gigabytes of data being created per hour by connected cars already and data being stored on premise, in data centres and in the cloud; the current data landscape is complicated and often disparate. Instead, a modern and centralised data architecture is required where data can be managed, stored, processed and analysed both at rest and in motion.
For manufacturers, this completely simplifies the process and allows them to concentrate on the insights generated from the data, rather than the headache of managing and controlling it. What’s more, the potential security risk of losing control or visibility of the data are also minimised.
A case in point
In the case of Nissan, the company was experiencing a variety of business challenges, namely the need for infrastructure capable of storing huge volumes of vehicle driving data and product quality data on a long-term basis. Also, it had a need for a Hadoop platform capable of deploying a variety of data cross-functionally.
Nissan turned to Hadoop as the solution for its Big Data problem, relying on Hortonworks Data Platform (HDP). The open source model appealed to the company due to the large numbers of engineering talent in the market and the flexibility to pivot if circumstances changed down the road. The deployment has introduced a data lake capable of storing all types of company data. The demand for and deployment of data continues to increase significantly.
While some people are hesitant to trust machine learning, AI, and autonomous vehicles, these trends can be more effective than relying on humans. In the same way that Netflix, Amazon, and Spotify are able to make highly refined recommendations based on the sheer quantity of actionable insights they’ve gathered from data, self-driving cars will continue to become smarter the more driving data available.
Autonomous vehicles aren’t perfect yet, but the trends suggest they will continue to improve. When they reach the point of being publicly available there is little doubt that they’ll be significantly safer and smarter than relying on human drivers.
Much like when Henry Ford transformed the automotive industry on the principle that, ‘if I had asked people what they wanted, they would have said faster horses’, the industry is on the brink of the next revolution. Data is the horsepower at the centre of this.
The author of this blog is Michael Ger, general manager, industrial manufacturing and automotive solutions at Hortonworks