How Big Data is driving the auto industry
Times are definitely changing when it comes to the transportation industry. Powered by big data, AI and the internet of things, our transportation system is poised to become more efficient and smarter [1]. Some hot topics include self-driving cars, autonomous trucks, connected cars, remote controlled cargo ships, etc. Even though we haven’t yet achieved a widespread adoption of some of these technologies, they have already been tested, prototyped and it’s clear where the future of the auto industry is.
Connected cars, equipped with access to the internet’s wireless ecosystem, can share this access with devices inside and outside the car. These devices work together through APIs and other interfaces such that concepts like monitoring, governance and administration apply to the entire ecosystem. It is powered by Big Data, with self-learning and self-reporting. These cars have numerous benefits like mobility management. This includes functionalities like optimal route finding, fuel consumption optimization and live traffic information. Additionally, it can predict vehicle maintenance, enhance safety and avoid collisions, enable a suite of entertainment options etc. Moreover, the car manufacturer can get data about the performance metrics of the car, insurance companies can learn the driving behavior, infotainment companies can understand the usage patterns, etc.
Given that the connected car can generate 1TB data/per day on average, the question is how to process such massive amount of data in real-time to predict any anomaly or fault before it happens [2]. This is being enabled by efficient data processing systems combined with the power of edge computing.
Self-driving cars employ a suite of technology, including GPS receivers, short-range wireless network interfaces, cameras, and an array of sensors, in an effort to uphold inter-vehicle communication. Vehicles interact with each other, but also communicate with sensors embedded in highway infrastructure and satellites, to help navigate streets, highways, and terrain, down to the nearest inch. It generates nearly 1 Gigabyte every second [3]. It uses all that data to know where to drive and how fast to drive. It can even detect a new cigarette butt thrown on the ground and it infers that a person might appear all of a sudden from behind a corner or car [4]. There are several other data sources, such as telematics or weather data, that can be also used to improve the driving experience [5].
Research has found that 81% [6] of car crashes are the result of human error. Whether it’s speeding, failure to comply with road laws, distracted driving, texting, drunk driving, or fatigue, there are clearly many factors that contribute to the number of accidents on the roads. Self-driving cars have even been called “more courteous and defensive than normal drivers” by Google’s director of the self-driving car program. Google’s fleet of autonomous vehicles are currently in testing, and eclipsed the 3-million-mile mark last year [7].
A major concern around all this is the security of the connected car. With these hundreds of OEM components, there are many opportunities for something to go wrong. In 2016, we saw how any Nissan Leaf could be hacked [8] from anywhere in the world as long as you had the VIN number of the car. This was due to an unsecured API in the HVAC component shipped by an OEM vendor. Not only device-level security must be addressed, but also data security during transmission and storage.
References:
1. https://www.forbes.com/sites/bernardmarr/2017/11/06/the-future-of-the-transport-industry-iot-big-data-ai-and-autonomous-vehicles/#4ecff0b21137
2. https://hortonworks.com/blog/how-big-data-is-paving-the-way-for-the-connected-car/
3. http://www.kurzweilai.net/googles-self-driving-car-gathers-nearly-1-gbsec
4. https://datafloq.com/read/self-driving-cars-create-2-petabytes-data-annually/172
5. https://infocus.dellemc.com/william_schmarzo/big-data-in-automotive-and-machinery-using-analytics-to-deliver-better-products-and-a-more-fulfilling-driver-experience/?utm_source=datafloq&utm_medium=ref&utm_campaign=datafloq
6. https://www.autoinsurancecenter.com/top-20-pros-and-cons-associated-with-self-driving-cars.htm
7. https://www.recode.net/2017/5/10/15605054/alphabet-waymo-self-driving-3-million-miles
8. https://jalopnik.com/how-the-nissan-leaf-can-be-hacked-via-web-browser-from-1761044716
Users who have LIKED this post:
4 comments on “How Big Data is driving the auto industry”
Comments are closed.
Hi Prachi! Great research on big data in the automative industry!
I find it really interesting how these self-driving cars are more “courteous” than human drivers. However, self-driving cars are only courteous because they are trained to avoid and react to events based on their training data (although they are given a lot of data). They are inherently biased; they only know what they have been trained to react to, whereas humans are exposed to a lot of situations and are constantly evolving without having to be programmed by an engineer.
According to the statistic you provided, 81% of car accidents occur due to human error. But do you think self-driving cars will complete eradicate this? What about computational errors that could occur while the self-driving car is driving (i.e. being unable to classify a possible roadmap and identifying the correct behavior)?
Even though I somehow tend to agree more on your point on biased reactions of self-driving cars after Craig Martell’s speech on AI and Machine Learning, I still believe that autonomous vehicles will be able to solve numerous problems humans are facing when it comes to transportation.
For sure there is still a long journey, as Prachi pointed out on security issues, but apart from reducing considerably the amount of accidents related to human drivers, I think self-driving cars will be able to solve traffic issues for example. I believe that commutes could be reduced dramatically if data from autonomous vehicles could be connected to traffic offices and therefore, traffic lights managed in a dynamic way, which would lead to a more efficient traffic management.
Hi Prachi,
Great post about how the automotive industry relies on large amounts of data in order to make decisions quickly. I have no doubt that our society is moving towards a world where every car on the road is autonomous. However, what is going to be extremely interesting is the transition when the road is half self driving and half human controlled. The interactions on the road will be extremely hard to predict as humans will have to make decisions not knowing whether the cars around them are driving themselves or driven by a human. Data will most likely get even bigger at this point as it will be a lot for a car to take in when they can’t seamlessly interact with just self driving cars, but they also can’t assume that every car on the road is driven by a human. The roads in this situation may become dangerous; however, this period of time will probably be short lived as self driving cars are going to become popular at an exponential rate.
This is an interesting article: https://www.wired.com/story/when-will-self-driving-cars-ready/
A comment on what the future is going to look like: Can people give a reaction on the question how they see the future. Do you think there will be to much people attached to their “own little space” or do you think they are willing to give that up for a cheaper way of transport? Do you think the future of cars is underground or in the air? Im really curious on peoples view.
Thanks in advance