Big data and the challenges in the car industry

By | 4 minute read | January 12, 2017

Auto show image

Another day at NAIAS 2017, and IBM have joined Intel and KPMG to talk big data and self-driving cars.

Gary Silberg, Partner and Automotive Leader for KPMG, started things off.

Six years ago, before it was chic, before it was cool, I wrote a paper: Self-driving cars: the next revolution. My colleagues thought I was crazy, but it got huge attention in the press and auto industry. I followed that up with: Self-driving cars: are we ready? This paper focused on the consumer side of self-driving cars. We found that the consumers were open to the idea of autonomous vehicles. We asked them to rank their trust, and customers trusted premium brands more than traditional brands if the vehicle was autonomous. So we followed that up and asked what if, instead of an auto company, it was a tech company: Apple, Samsung, Microsoft, etc. To our surprise the public said they would trust a technology company more than a traditional or premium car manufacturer.

We have three other papers: I see I think I drive (I learn) – focusing on Deep learning and AI. I promise you deep learning is the most important technical advancement for autonomous vehicles. We also have: Clockspeed, capable procurement – it explores how large OEMs think about their procurement strategy going forward. Auto companies and tech companies need to work together; the technology is instrumental to give the OEM the advantage of being first. How can the two can work together and smarter? And next up we have: Your connected car is talking, who’s listening? – a paper about the connected world and hackers, how they come together and what you need to do about it.

We were then introduced to Kathy Winter, VP and GM, Automated Driving Solutions Division, Intel.

Beside big data why are we interested in autonomous driving? Well $500 billion could be saved in traffic accidents and the cost to society. And $507 billion could be saved in productivity gains. Think about if you could get that 90 mins back each day? Better traffic patterns and logistics lead to big savings.

The overall auto model is changing. We don’t just want driving anymore, we want fleet, no need to own the car. And then there’s the possibilities for the media and the content that can be delivered into the vehicle.

But with a 10 x increase in data from a vehicle by 2025, how do we manage that data? What can we do with it? Something has to change.

Do we really think self-driving will happen?

For sure. When you look at it, there are scalable developer kits readily available. We can speed up the innovation in automotive space, and that’s good for everybody. An autonomous vehicle is much safer than a speeding young teen driver, or an elderly person with slow reflexes.

5G will be crucial to get the speed of data up and back. A data center to use deep learning to constantly update fleets.

Following on from that, Sachin Lulla, Global Watson IoT AutoLAB Leader, at IBM, took to the NAIAS stage.

Sachin Lulla and the future of the self-driving car

Sachin Lulla and the future of the self-driving car

AI and the self-driving car

A lot of people are concerned that AI (Artificial Intelligence) will replace human jobs. At IBM we see AI as augmented intelligence not artificial, because if you lose the human element you have lost the game. It’s all about improving human lives.

Today I wanted to cover the following:

  1. Demystifying AI and showing what AI can do
  2. Framing the context of the opportunity the auto industry has, plus big data, and how we are changing our relationships with vehicles
  3. To share some examples from IBM Watson IoT AutoLAB

Take a look at what we have done from a consumer perspective:

Olli took less than 3 months to build, has a personality, and learns you preferences. We want to redefine the experience and relationship that we will have with future vehicles. You can read more about Olli on our IoT blog.

People are clearly expecting these types of services as quickly as possible. The car needs to start anticipating their needs and start taking care of the passengers.

Currently the in-car experience is based on features and functions, rather than an intelligent experience. We’ve made some progress, but is that the best we can do? The Olli example shows that it can experience and learn; it transforms how we drive, how we travel.

We travel on average 46 minutes a day. That’s a lot of time in car, and you can feel captive. Our goal is to use cognitive systems to transform how people use their time.

Soon every vehicle will be connected. 10 million by 2020, according to Business Insider, or 250 million by 2020, says Gartner. The data footprint will be huge. Data will become the new fuel, and cognitive systems like Watson will be the second engine to turn this data into value.

Personalisation is key to the future: personalisation leads to monetisation

We created OnDash – the world’s first, and smartest co-pilot, a commerce and cognitive platform powered by Watson. It uses data from your vehicle, and knowledge from interactions within the vehicle, and turns that information into personalised commerce experiences. It’s white label, so we can partner with any EOM, not just General Motors, as we announced at World of Watson.

And at CES we launched the world’s first cognitive infotainment platform. It’s Watson and Panasonic, serving a world-class customer experience, and taking advantage of all data opportunities.