“If you want to know the future of fleet, you need to look outside the fleet industry.” These words were spoken to me several decades ago by Jim Noonan, then-senior VP at PHH, a past fleet management company now owned by Element Fleet Management. I never forgot this insight and, today, I light heartedly identify it as the Noonan Axiom.
Jim’s contention was that the fleet industry is not an incubator of new inventions, but, as an industry, we have a long history of being early adopters of new technologies and business practices that originate in other business segments.
With this observation as a prelude, let’s take a look at other industries that today are on the cusp of disruptive technological change to see if these technologies may be transferrable to the fleet industry. One industry in the vanguard of dramatic changes is the healthcare industry driven by the adoption of cognitive computing platforms.
Most of us are probably most familiar with IBM Watson, an artificial intelligence (AI) cognitive computing platform that in early 2011defeated two past grand-champions on the TV quiz show, Jeopardy. This was a seminal event in the history of computing and a harbinger of changes to come.
Today, there are more than 20 cognitive computing systems in the market, such as Microsoft Azure, IBM Watson, Google Cloud Prediction API, Wipro HOLMES, and Infosys Mana. These systems provide the foundation to create enterprise-specific AI platforms that tech companies hope to deploy in many industry verticals, such as health care, financial services, manufacturing, to name just a few. I predict enterprise AI platforms will also find a home in the fleet management industry.
At a fundamental level, the fleet management industry is an aggregator of data upon which it executes actions designed to optimize vehicle asset lifecycle – from acquisition to disposal – and to fine-tune operational efficiencies to maximize employee productivity. A cognitive computing platform would thrive in this type of data-rich environment.
For example, the IBM artificial intelligence system can analyze high volumes of data, understands complex questions posed to it in natural language, and proposes evidence-based answers using computer-generated natural language. Think of a “Siri for Fleet.” Watson employs what is known as “machine learning,” and its subset “deep learning,” which is a collection of algorithms that enable a computer to learn complicated concepts by building them out of simpler ones.
Using Watson technologies, IBM launched Watson Health, a health care-dedicated division in 2015 with the goal to turn it into a $15 billion business. Today, pharma companies are utilizing Watson to collect and analyze enormous genomic data sets to identify possible treatments that are genetically tailored to an individual’s genome. One patient’s electronic health record holds an average of 400 gigabytes of information. If you add a patient’s genetic information this increases to 6 terabytes of health data.
In comparison, a connected car sends 25 gigabytes of data to the cloud every hour, while an autonomous vehicle will generate 4,000 gigabytes of data per day. Big Data analytics will increase fleet management granularity to custom manage each vehicle to its unique parameters, to where potentially each unit could have its own P&L.
IBM has opened its Watson platform to allow others to develop their own programs and more than 7,000 applications have been built through the Watson Ecosystem. How might fleet management utilize IBM Watson? Here are several possible scenarios.
Remarketing Services: In 2015, IBM acquired Merge Healthcare for $1 billion, which has a database of 30 billion medical images. The goal is to use deep learning to spot patterns in data for future prognosis and treatment. Could similar outcomes be achieved in vehicle remarketing? For years, vehicles to be remarketed have been photographed at a variety of angles for use in online sales.
Could an AI-based deep learning of this vast database of images, integrated with historical sale data, yield more precise reconditioning strategies and determine the specific dollar impact on resale based on vehicle condition? Or, could an IBM Watson system remotely rep a vehicle auction using up-to-the minute resale market data in ascertaining what is a fair market value for a vehicle? Conceivably, an IBM Watson-type system could arbitrage geographic price differences for particular makes and models of vehicles. Perhaps, at some future date, this could be a machine-to-machine transaction governed by specified policy parameters.
Maintenance Management: One application could be to augment the maintenance departments at major fleet management companies. Currently, maintenance advisors work with service providers to ascertain if certain repairs are necessary and authorize them within certain cost guidelines. Could this process be automated using IBM Watson with its computer-generated natural speech capabilities? Here’s a wild thought: As machine learning cognitive capabilities advance, at a future date, could an IBM Watson system become eligible to be ASE-certified?
Continuous Driver Risk Profiling: Could a cognitive computing system identify and continually assess the indicators that make an employee a high-risk driver? Perhaps this can be integrated with additional data collected from employee-worn wearable sensors and computers.
While cognitive computing systems are still in their infancy, the technology is advancing quickly and its applications will be far-reaching. One of the newer initiatives by IBM Watson team is in the area of connected vehicles and the Internet of Things, which promises to further change traditional fleet management practices.
Let me know what you think.