MV Transportation, the largest private provider of paratransit services and the largest privately owned transportation contracting firm in the United States, has more than 20,000 employees, primarily drivers and maintenance personnel. But some of its most important employees may be the handful of data scientists that Jim Haring, MV chief information officer, has recruited to give MV a leg up over its competition.
DXC Technology interviewed Haring to capture his insights on how emerging technologies like artificial intelligence (AI), machine learning, virtual reality (VR) and autonomous vehicles will play into the industry. He also discussed the ways in which disruptive newcomers such as Uber and Lyft might one day integrate into the broader ecosystem of paratransit and public transit mobility as a service.
Tell us a little about MV Transportation.
The company was born out of the paratransit service industry, which you might even say we started. The husband and wife team of Feysan and Alex Lodde, who still own the company today, started MV in San Francisco in the 1970s, providing transportation to the elderly and people with disabilities. The passage of the Americans with Disabilities Act in 1990 catapulted the paratransit industry because everybody had to comply with new regulations. Today, paratransit is still nearly 50 percent of the company’s business; 30 percent is fixed route, which is major city buses; the rest consists of school buses, and corporate, university and airport shuttles. Across all lines of business, we are transporting over 100 million passengers each year with a fleet of more than 11,000 vehicles.
What are some of the biggest challenges MV faces in the marketplace?
MV bids for contracts with local governments, which means the bids are publicly available to everyone, including our competitors. Margins in this industry are low, and primarily we compete on the labor costs for drivers and mechanics. We’re all competing for the same labor pool, so if we win a contract, a driver or mechanic who worked for the previous incumbent just takes off his or her badge and switches to MV. But it’s the same person.
My big challenge right now is, how do I digitally transform my company? How do I use AI to make my humans more efficient or smarter than my competition? We’ve leaned on a strategy of human-enabling technologies. I can empower those people through data, through AI and machine learning. That’s the biggest bet we’ve made: that data science will become the intellectual property of my company and will help me differentiate.
When I first joined MV Transportation, we had a 10-person application development team that had been building a piece of software for 5 years with very little monetization. They were burning through a large budget each year, not using current tools; everything was manually intensive with very little automation. Dallas, where we’re based, is a competitive market. It’s very expensive to attract the right talent. My question was, “Why is a transportation company building software, anyway?” DXC provided an influx of application development talent and an influx of current-generation tools to enable MV to complete the project and envision new ways to address future projects. That’s my goal: to use technology that will be key to MV’s future vision and to invest in projects that add to our intellectual property in our unique industry.
We see trends that focus on worker productivity having big impacts in the coming years. The pervasive use of AI and machine learning in business — coupled with collaborative systems — is paying off. What are the key trends you see happening in the industry over the next 3 to 5 years?
I agree. In our industry we will focus on: as-a-service, safety and security. My customer is the transit agency. The challenge it faces is that ridership is generally decreasing and there is a disruption in the marketplace from the Lyfts and Ubers. We are searching for creative solutions to integrate all those modes of transportation into a mobility-as-a-service offering. Everybody’s doing some kind of pilot, looking for a collaborative, broker model that gets more passengers on more shared rides.
Just like everything else, the goal is to better use data to make the best use of resources. In this case, how do we best utilize existing capacity and integrate the different fleets. I think analytics technology will ensure that mobility as a service happens in the next 3 to 5 years.
The second area is safety, which is one of the largest cost factors for the company. Just like the average driver, our bus operators have a safety record that determines what rates our insurance company charges. Today, we have cameras in 80 percent of our vehicles. The cameras are triggered when certain events occur, such as hard braking, rapid acceleration or sudden turning. We use AI to help us evaluate these actions. Drivers are scored every day on their performance, and people who reach a certain risk factor can be taken off the road and sent for retraining or even be fired. After 18 months, MV’s insurance rates have dropped by more than $20 million.
And security should be top of mind for any business. We don’t want our customers’ information compromised; we don’t want to end up on the front page somewhere; and we don’t want the business risk that comes from not paying enough attention to all forms of security. We’ve done extensive analysis on ways to prevent hacking, ransomware threats and other security risks. DXC brought us the most valuable information on these areas and showed up with meaningful expertise and advice. DXC’s security experts provided forensic analysis and shaped our thinking on how to ward off and prevent security issues. We’ve learned a lot about how to reduce risk and increase cyber resilience, which has now shaped our efforts for all future projects. DXC has some great options for multi-layered defenses that proactively detect unknown threats and protect critical enterprise information.
With more and more uses for AI in many industries, what are some examples of how AI and machine learning are paying dividends at MV?
We are trying to take advantage of any technology that can better run our business or help us more efficiently monetize the work we do. In addition to using it for our driver-tracking program, we’ve started a predictive maintenance program. We take 10 years of work order history and analyze it, using standard tools in the Microsoft Azure cloud. We ask all kinds of questions. Why do we change the oil at 6,000 miles? Does route play a factor? Or weather? Or heat? We also have two data ports on the vehicles that send real-time telematics back to the cloud. We get information on things like oil temperature and battery voltage. Then we use machine learning in the Azure public cloud to make predictions on which vehicles are most likely to fail. This enables us to adjust our maintenance schedules and fix a bus before it fails. And we’re not building anything new; we simply consume public cloud services.
Another area we are looking at is AI assistants for our employee support. It’s a nice option to further automate routine IT processes and support. We recently migrated our internal help desk to a DXC service desk facility in Manila [Philippines], which was probably one of the first offshore moves in our industry.
At MV, we had never done anything like this before. In terms of impact, it was projected to be a big culture change, but it’s had a big, very positive impact. Employees were concerned about quality of service and language issues. IT was looking to take advantage of offshore talent to free up budget for our core projects. Today, the employees enjoy better support, and the language issues they had expected never materialized. It’s hard to tell where in the world support is coming from, and they now believe we can provide other quality IT services from anywhere. Adding AI software agents could augment and improve other aspects of their work lives.
A number of organizations that are short on experienced labor are implementing augmented reality or virtual reality to maximize or extend the knowledge of their more experienced employees. Are you doing anything with virtual, augmented or mixed reality?
We have a pilot project with Microsoft’s HoloLens 2. The biggest use case would be remote assist, for example, where a mechanic in Dallas could help guide someone in a remote location — but adoption by the mechanics is going to be a tough challenge. It should show some promise after the next year or two.
Can you describe the business results that you’ve achieved through these digital transformation efforts?
We think we’re probably 6 months to a year ahead of the competition. Not many companies are getting into predictive and prescriptive analysis. And we’ve had the four biggest wins in the industry this year.
Our automotive industry clients are racing to develop self-driving cars. What role do you think autonomous vehicles (AV) will play in your industry, and how is MV preparing?
To us, it’s just another mode of transportation; the only difference is the operator. But the vehicle still needs to be maintained and still be part of the overall ecosystem of routes and schedules and customer service. We already have three or four AV pilot projects, which tend to be closed-loop, point-to-point routes. For example, MV is a shuttle partner in the National Renewable Energy Lab’s autonomous electric vehicle employee shuttle service on its Colorado campus.
But we think it’s going to be many years before we have autonomous vehicles in major cities. In the meantime, we’re preparing for the big data problem associated with autonomous vehicles, because there will be 100-fold more data coming off each bus. The company that can best collect, analyze and get the most insight from that data is the company that is going to win.
What are the latest trends in paratransit, and how are you using technology to improve the lives of elderly riders and people with disabilities?
Paratransit is a unique business. Tech adoption is very low. Some riders might be in a wheelchair, but others might be ambulatory and have a sight or hearing impairment. We launched an app that enables paratransit riders to schedule pickups online, but the adoption rate is only in the 20 to 30 percent range. Many riders still want to call on Tuesday to request a ride on Thursday.
In the Seattle area, we have a program where we come to the paratransit rider’s house and try to determine the best pathway to get the rider to the public transportation system, where a fixed-bus ride costs $6, compared to between $60 and $90 for a handicapped-equipped van. We come to your house and train you. We collect all kinds of data — the best route to the bus, where the curb cuts are, streets to cross and how many bus stops it takes for the rider to reach the destination.
We’re also looking to collect all that data digitally and store it in a common data source, so we never have to go back to that street corner, for example, to determine whether there’s a curb cut for a wheelchair. This will be more efficient for us and should allow riders to discover other routes in addition to their regular trips. As riders are successful with the app, they will learn to rely on it more and more.