Big data in the broadly understood transportation industry is nothing new: airlines are known to leverage large quantities of data in their pricing strategies, bike-sharing providers bank on data to manage geographical distribution of their assets, and traffic applications such as Waze use real-time data to optimize navigation. Ford takes its commitment to big data one step further: it wants to understand driver usage patterns to create better transportation for everyone.
The automaker recently opened a lab in Silicon Valley with exactly that goal in mind: to improve its automobiles, their integration, the driving experience and the transportation system at large with big data. Three experiments were launched in parallel:
- Fuel consumption and safety: Four million gasoline-fueled cars were equipped with in-car sensors in order to deliver a whole new level of real-time insights to engineers on factors affecting fuel consumption and safety. This project aims to analyze a vast array of inputs, from weather conditions, through car’s maintenance state to individual driving style, and translate those inputs to a better driver experience. For example, the brake system is measured along 25 separate dimensions – how hard the driver pushes them, at what speed, during what accompanying conditions – to better predict failure.
- Power usage in electric cars: Vehicles from Ford’s newest electric cars line, the Energi, have been equipped with an impressive number of sensors – 74, to be precise – including sonars, cameras, radars, accelerometers, temperature and rain measuring devices. Additionally, information is gathered on recharge plug-in times, locations, and lengths, helping Ford understand patterns of consumer behavior in this relatively new field of car charging. As a result, every hour Ford finds itself on the receiving end of approximately 25 gigabytes of data.
- Behavior of employee fleets: Ford analyzed driving patterns of employees that were part of large fleets in companies that utilize mobile salesforces (and drive Fords). This experiment focused mostly on individual usage patterns such as driving routes and car utilization.
Creating value for shareholders, customers, and the society at large
Insights from the early stages of the experiments are already proving useful. For example, Ford plans to utilize the safety data from experiment (1) to better understand the influence of individual driving patterns on number and severity of incidents on the road. The company believes that this will lead to more accurate insurance quotes, benefitting both the insurance companies and the society at large (incentivizing safe driving behavior). Ultimately of course, Ford’s shareholders stand to gain because those insights won’t be given away for free… Experiment (2) has primarily the customer focus in mind: together with analyzing recharge data, Ford is developing a driver-facing app that will feed the car owner real-time information about their vehicle, developing a personalized recharge plan. As a result, the owners are expected to decrease both their total ownership cost and the frequency of occurrences when their car runs out of battery. Ultimately of course, Ford’s shareholders stand to gain because this experiment tackles the key obstacle prohibiting electric car proliferation. Experiment (3) generated many usage insights that could be turned to the benefit of the fleet owners. For example, traveling employees left their vehicles unused at the airport for days and 70% of trips took place during weekdays, opening possibilities for car-sharing schemes aimed at decreasing pollution. The experiment also provided a percentage split into four separate driving categories alongside a two dimensional matrix, each of which had different risks, fuel consumption and asset utilization characteristics associated with it: city block driving versus freeway driving; non-rush hour driving versus rush hour driving. Such insights will allow fleet owners to optimize trip schedules and modes of transportation employed. Ultimately of course, Ford’s shareholders stand to gain if Ford manages to position itself ahead of the car-sharing game.
Challenges on the road ahead
Creating value from big data analysis is often the easier half of the challenge; capturing it is much more difficult. Most of the insights identified by Ford present long lead times from generation to implementation, delaying the value capture moment. In many cases, Ford is also banking on being the early mover to capture network effects (electric car app, car-sharing schemes), making the value capture not only delayed but also heavily uncertain. Another challenge lies with experiment themselves: are the samples unbiased and representative? Are the insights generated statistically sound given the vast array of factors influencing the dependent variables, of which still only a handful can be measured? Can Ford find the right balance between conducting the experiments long enough to create powerful insights, but short enough for competitiors to not follow suit? And more broadly, is following the “Apple of the automotive industry” strategy, i.e., deriving synergies from the integration and intelligence between hardware and software, right for Ford?