Beyond On-Time Performance: Using Transit Data to Understand What's Driving Results

Earlier this year at the ThinkTransit conference, TransTrack’s Yuvaraj Mani presented a session titled “Beyond the Rearview Mirror: Leveraging CAD/AVL Data & Real-Time Insights to Drive Transit Innovation.”

He opened with a simple but important question:

What does your CAD/AVL system actually do for you?

For most transit agencies, the answer comes down to three things:

• Vehicle tracking
• On-time performance (OTP) reporting
• Incident management

And in nearly every conversation with transit leadership, one theme consistently comes up: On-time performance is the metric every agency focuses on.

But here’s the challenge. OTP is an outcome. It tells you what happened—but not why. When a route is consistently late, the instinct is to flag it as a performance issue.

But what if the schedule itself was never realistic? What if the running times were misaligned from the start?

That’s the gap many agencies are facing today.

They’re measuring results without examining the inputs driving them.

The reality is, your CAD/AVL system is already capturing the data needed to understand:

• Running time variances
• Dwell time patterns
• Headway consistency

The opportunity lies in actually using that data to inform decisions.

At Guelph Transit, planners used the ViewPoint Running Times module to analyze run times and time bands, which provides up-to-date information in a simple and intuitive manner. This reduced the time needed to implement a route change from 30 to 7 days.

Instead of relying on ride-alongs, stopwatch timing, and weeks of back-and-forth review, Guelph was able to make confident, evidence-based schedule adjustments using the data they already had.

The result? A 77% reduction in route change cycle time, giving planners weeks back to focus on service quality.

With ML-powered running time recommendations, agencies can move beyond reporting and make proactive schedule adjustments before performance issues become chronic.

Instead of reacting to performance issues, they can prevent them.

The shift from monitoring outcomes to addressing root causes was a key theme throughout Yuvaraj’s session. And based on the conversations in the room, it’s clear the industry is ready for it.

Your CAD/AVL system already knows more than you think. The question is whether you're turning that data into action.