Over the years, the Big Data concept has evolved into an all knowing, all-encompassing technological behemoth, devouring everything that lies before it. For the uninitiated, Big Data is simply a notion to express the very large amounts of data that is being generated in our age of the Internet in the form of text files, images, videos, web pages, mobile data and so on. The amount of data in Big Data is not defined, but rather an abstract value that we regard as so huge, that traditional data storage technologies are not considered efficient enough in collecting, storing, and analyzing the said data. Thus, they require the need for a specialized set of tools, software platforms, and databases that we collectively call Big Data Technologies. These technologies are often used by corporations and institutions of all shapes, sizes and forms from super-market chains to space organizations, who try their hand at finding out whatever hidden meaning they can dig out from the immense amounts of data generated by their businesses.
One such opportunity in Big Data which we will be looking into is the automobile industry, and more specifically Telematics. Telematics comes from the French word tlmatique (which is a combination of two other French words telecommunications and informatique), meaning transfer of information over telecommunications. Gartner, a leading IT analyst, defines telematics as such the use of wireless devices and black box technologies to transmit data in real time back to an organization, typically used in the context of automobiles, whereby installed or after-factory boxes collect and transmit data on vehicle use, maintenance requirements or automotive servicing.
In simpler terms, telematics is gathering and analysing vehicle data, i.e, data generated by devices installed in an automobile. The data is then wirelessly streamed, usually via existing cellular technologies, to a backend server, where data-mining and data-crunching takes place. The telematics device records all sorts of data from the vehicle, ranging from GPS coordinates, various vehicle metrics, such as engine performance, mileage, water temperate, speed, steering angles, acceleration, and braking frequencies, and also data from the vehicles on-board entertainment console. With such complex and varied data in hand, one could get insight about a vehicles performance, driver behaviour, logistical patterns and so on and so forth. Such insights over potentially hundreds of thousands of vehicles over time will no doubt prove beneficial to a wide array of companies in insurance, logistics, car entertainment, policy makers, safety auditors and of course, vehicle manufacturers.
Devices also vary between manufacturers. Most modern cars come with one built-in device. For others, a 3rd party device is fitted somewhere beneath the dashboard (usually under the steering column). A more recent advancement in this field is the introduction of Googles Android Auto and Apples CarPlay, where a user is simply required to plug his/her smartphone into the cars USB port, after which the phone doubles up as a telematics device. The obvious advantage of this is that it does away with the necessity of having to install another device and also, a driver will now be able to use his/her phone using inbuilt car controls. A company considering to jump into the telematics bandwagon may now only be required to develop an app for the concerned platform and have the user run it when behind the wheels.
Telematics and Big Data
As one can imagine, the amount of vehicle data expected to be gathered for mining, crunching and analysis in Telematics is huge, considering the vast array of metrics that will be streamed from thousands or even millions of vehicles over time. An IBM whitepaper reveals that the volume of data collected from 26 million connected cars in 2013 was more than 480 TB and this number is expected to jump to 11.1 PB by 2020. It also states that some hybrid vehicles can generate up to 20 GB of data in just one hour! Further, InsuranceTech made an observation that within a year of enrolling 1,000 average drivers on a UBI (Usage-based insurance) program, an insurance carrier must accommodate the transmission and storage of over 190 million data points. That is a crazy amount of data generated and streamed to a data centre at an astonishing rate, sometimes even on a per second basis.
Getting the data into a database is one thing and analysing the staggering amount of data another matter altogether. Traditional data warehousing technologies are ill equipped to handle the huge amounts of data at the mentioned frequencies. Its architecture just cannot cope with the volume, velocity, and variety of data. Companies are therefore looking at Big Data alternatives in popular distributed computation frameworks such as Apache Hadoop and Apache Spark. These platforms, along with others in the Big Data ecosystem, are enabling developers and IT companies in processing unprecedented amounts of data to gain never before seen insights, patterns, and hidden opportunities in Telematics and other such enterprises.
Untapped Telematics Opportunity in India
The automobile industry in India is getting bigger. Although India stood at a lowly 160th position in a list of countries by the number of road motor vehicles per 1000 inhabitants, with only 18 vehicles per 1000 persons (2011), the situation seems to be improving. According to SIAM (Society of Indian Automobile Manufacturers), the industry produced a total of 23,366,246 vehicles including passenger vehicles, commercial vehicles, three wheelers and two wheelers in April-March 2015 as against 21,500,165 in April-March 2014, registering a growth of 8.68 percent over the same period last year.
The table below details the domestic sales of vehicles in India over the years:
The key takeaways from the data are:
The sales of Passenger Vehicles grew by 3.90 percent in April-March 2015 over the same period last year. Within the Passenger Vehicles segment, Passenger Cars and Utility Vehicles grew by 4.99 percent and 5.30 percent respectively.
The overall Commercial Vehicles segment registered a de-growth of (-) 2.83 percent in April-March 2015 as compared to same period last year. Medium Heavy Commercial Vehicles (MHCVs) grew by 16.02 percent.
It goes without saying that with the number of Passenger and Commercial vehicles, especially the Medium Heavy Commercial variants, is steadily increasing over the years and with no widely known Telematics related endeavours taking shape, this particular brand of technology in our country is ripe for exploration. The prospect is especially enticing in a new and vibrant market, such as India, with a growing economy coupled with varied cultural and geographical landscapes presenting to those who dare to undertake this venture with a colourful challenge and incredible bounty.
Telematics Use Cases
Lets take a look at a few examples of how this technology is used in various industries.
The insurance companies and their customers are perhaps the greatest benefactors of this technology. Currently, automobile insurance premiums are given on the basis of the type and cost of a vehicle, and taking into account that a certain class of vehicle would draw a certain premium based on the case history pertaining to that category. The driver is not taken into consideration at all. With a telematics device on board, the insurance company can constantly and accurately monitor a drivers driving habits throughout and correctly reward them with lower premiums for safe driving and also charge higher for those who are more aggressive in their handling of their vehicles. Another advantage crops up when dealing with insurance frauds. Since the status of the automobile is monitored at all times, it gives the company all the details it needs to correctly judge if a case is genuine or otherwise.
Some nifty insurers provide their customers a near real-time interface to monitor their driving habits themselves, thereby allowing them to adjust their driving patterns on the go.
Popular snack food manufacturer Frito Lays made public their success story of using telematics in managing their army of delivery trucks. They claim that a supervisor can now monitor a fleet of 500 vehicles at a time when he could monitor only 50 earlier. They were also successful in reducing idling time down to 50% as well as insurance claims from anywhere between $1000 to $2000 per vehicle. DHL and many other delivery services are making use of this technology to better plan transport routes, manage fuel economy and gather so many more insights to enable them to run their fleets more efficiently.
Car entertainment presents a very unique and interesting set of opportunity in telematics. Every modern vehicle now comes equipped with a standard music system and other media peripherals. By gathering user preferences over time, one can predict a behavior pattern and subsequently suggest customized media for the drivers consumption. By using GPS data, a mechanism similar to an in-flight suggestion system could be developed for cars and vans, whereby the driver is alerted of popular tourist attractions, fuel depots, restaurants, ATMs etc. Although such systems already exist, using telematics, suggestions can be personalized from a vehicle owners perspective. Imagine an intelligent system that dynamically changes music depending on the style of driving or surrounding traffic. It could automatically play soulful or ambient music when it detects that the car is cruising on a highway or could switch to heavy death metal when it learns that the car has been idle right smack in the middle of a 4km long traffic jam for the past few hours (and thus aid in fueling the fury of the driver).
Similar to the example stated above for dealing with insurance frauds, investigators and safety auditors can now accurately study the data relating to the state of a vehicle that has met with an accident. They could learn the circumstances that led to a mishap and then possibly devise counter measures ensuring they dont happen again. Incidentally, they could advocate better vehicle safety features in a vehicle after studying the data at hand.
Studying the data generated by their vehicles, manufacturers can be better prepared to develop better variants and upgrades. With a host of data varying across matrices, they can look to improve anything from fuel efficiency and engine performance, to driver comfort, passenger ergonomics and environment friendly vehicles. It could also be beneficial when a customer brings in their vehicle for timely services as the service center would now be better positioned to run maintenance after studying how each component in the car has been performing thus far.
Government agencies can stand to gain knowledge on how the public drives on its roads, to find out which are the most congested and which arent. This could help them suitably plan budgets for maintenances or aid them in making an informed decision on alternate traffic routes. Further, they could plan on increasing the frequencies of public transport along the lines that see more traffic, thus, encouraging its usage more often. In times of emergencies, ambulances and related services can make use of the data to determine the quickest route to their destination that would have minimum blockades.
These are just some of the more obvious use cases in telematics. The information gained from this technology is also being used in developing autonomous self-driving cars, enabling manufacturers to come up with fine-tuned, fuel efficient, and environmentally friendly vehicles into the market. Here at HCL, we have already begun aiding our clients in setting up a backend infrastructure to regulate, monitor, clean, and understand the data provided by vehicles. As time progresses, we stand to understand more of what we have and help our customers and ourselves drive further.