How data-driven supply planning can transform your service.

At Via, we think supply planning should deliver three key outcomes: a high quality rider experience, happy drivers, and ultimately a reliable and efficient network.

A note before we dive in. We’ve heard that different teams use different terms to describe what we call supply planning. You might say “supply planning”, “shift planning”, “driver planning” or “run planning”. In any case, when we say “supply planning” in this article, we’re talking about the art and science of deciding how, where, and how many vehicles should be on the road at any given time throughout the day in order to efficiently serve ridership demands. 

So how do we get there? 

1. First, get local and understand your community.

The first step to running any successful service is understanding what your community needs before you begin planning. It’s critical to get to know your riders and drivers personally to create plans that serve their needs and provide opportunities for continuous feedback. 

Of course, this also means understanding the goals and constraints of specific services. For example, if you're running an ADA paratransit service, you'll need to consider factors like the service level requirements set by the ADA, labor agreements with your local union, and the types of wheelchair-accessible vehicles available in your fleet. These considerations of course look very different if you're running an on-demand microtransit service serving commuters.

What do we do at Via to get to know the community?

• Regular driver “office hours” and “town halls”
• Onboarding programs
• Local fleet management and vehicle maintenance
• Ongoing depot presence
• Ride ratings and in-app rider surveys
• Understanding of traffic patterns, seasonality, and weather dynamics that affect the community

What might you learn about your community of drivers and riders that would materially impact how you plan?

We always learn a few things! For example, at a recent launch, we anticipated a high number of cash-preferred riders needing to visit the utilities office to pay their bills during the first week of the month, so we accounted for the office’s operating hours in our planning. At another launch in a popular ski town, we understood the difference between peak ski season and the off-season and used that information to our advantage in planning.

When you understand the community, you can put in place key parameters you need to consider in service planning and optimize with technology. Keep reading to see how it works!  

2. Create the best service schedules by using data to understand demand and optimize supply. 

Once we understand the community and service needs, we can do some really interesting things with technology. 

Instead of putting pen to paper each month and creating driver schedules based on what we think demand looks like, Via’s Supply Studio—a daily automatic shift optimizer—leverages AI to forecast supply needs based on historical demand data and models from other similar services. 

What does that mean? It means we know exactly how many vehicles you’ll need on the road next Tuesday at 10:00am.  

The supply algorithm then translates your predicted demand into your required vehicle supply needed to meet it—down to the hour. 

But drivers can’t operate without breaks, vacation, or changes in vehicles. Many of our partners also have complex agreements with contracted operators or labor unions that dictate when and for how long drivers can work. So, we bring these two data streams (supply and demand) together to create optimized driver shifts that balance service efficiency and driver needs and preferences. Have more specific service needs to meet?

Customize preferences to meet your service goals.
Different services have different goals. All of which are important to consider when optimizing. For example, paratransit services need to provide excellent service while meeting ADA requirements, whereas microtransit services have goals to efficiently provide coverage in low density areas.

Via's Supply Studio provides several options for plans and allows us to choose based on your goals (even if they change throughout the year!).

3. The final step? Staying human—we serve happy drivers, so they can serve happy riders.

Thanks to the data-driven, flexible optimization process, Via drivers see shifts broken down by day and can choose to work the times that best suit them if this is possible given labor requirements of the service. This "shift claim" is integrated with the Via Driver App and is based on a balance of driver preference and labor requirements.

Integrated driver performance data also allows for additional positive reinforcement—reliable drivers gain preferential shift claims, helping ensure low callouts!

The bottom line?

Agencies can struggle with limited access to demand data, making it difficult to optimize for supply and difficult to frequently update schedules to match the optimal level of supply. 

At Via we use historical data from your service as well as hundreds of services like yours to power AI predictions of when and where vehicles are needed. This level of granularity helps us produce the best possible schedule. And, if we are operating your service, we can change schedules quickly in response to real-time data!

This more nuanced, flexible, and data-driven delivery of service not only improves network efficiency, but at the end of the day improves the experience of your people—your drivers and community. 

Want to learn more? Get in touch.