Transportation agencies are under increasing pressure to improve traffic operations while working with limited staff, budgets, and field resources. To make the most of their time, engineers need a way to quickly identify problem locations, prioritize improvements, and measure results across an entire network.  

Probe–based Signal Analytics uses vehicle trajectory data derived from anonymous GPS points to measure how vehicles move through signalized intersections and corridors. By providing continuous visibility into real-world traffic performance, it helps to identify problem locations, prioritize retiming work, and measure results without deploying permanent or temporary detection or running manual counts at every intersection. 

When to Use Signal Analytics 

Signal Analytics is especially useful when: 

• You need to triage a large network but don’t have staff for intersection‑by‑intersection reviews 

• Complaints are coming in, but it’s not obvious which intersection is causing the problem 

• You want to evaluate retiming results without running new field studies 

• Detector coverage is inconsistent, unreliable, or nonexistent 

• You need network level visibility before deciding where to invest limited time and resources.  

What Inputs You Need Before Beginning? 

You don’t need signal timing files or detector data to get value. At minimum you’ll need: 

• A list of signalized intersections (or a corridor) 

• A clearly defined time window (e.g., specific dates, weekday AM peak, PM peak) 

• One or two performance questions (e.g. Where is delay occurring? How well is progression working? Are queues spilling back?) 

That’s it.

What Probe Based Signal Analytics Measures  

Probe based analytics focuses on what vehicles experienced, not what was programmed. 

Key measures include: 

• Arrivals on Green (AOG): Did vehicles make it through without stopping? 

• Control Delay: How much extra time did vehicles spend because of the signal? 

• Excessive Delay: How many vehicles waited longer than 3 minutes due to having to stop more than once? 

• Travel Time & Reliability: What is the experience of single vehicles through the corridor from end-to-end and how does this compare to typical experiences? 

• Turn Ratios: How many vehicles did INRIX observed making left, through, and right movements? 

These measures are derived from real vehicle trajectories, aggregated at the movement, approach, intersection, and corridor levels.

Step-by-step: How City Engineers Actually Use This 

Step 1: Start Wide (Network View) 

Begin by scanning the network to identify: 

• Intersections with consistently high delay 

• Movements with recurring excessive delay 

• Corridors with poor travel time reliability 

This helps eliminate the common “where do we even start?” problem. 

Step 2: Narrow to a Corridor or Problem Area 

Next, focus on: 

• A corridor that is generating complaints 

• A cluster of intersections showing similar performance issues 

Define the corridor using observed turning movements rather than simply relying on arterial names. This ensures the analysis reflects how people actually travel through the network. 

Step 3: Identify the Problem Locations 

Within the corridor: 

• Rank intersections by delay and excessive delay 

• Look for repeatable patterns by time‑of‑day 

• Ignore one‑off spikes caused by incidents, special events, or construction 

Most corridors have two or three intersections, doing most of the damage. 

Step 4: Decide What to Fix First 

Use the metrics to guide effort: 

• High excessive delay counts → likely green time, demand or storage issues 

• Low arrivals on green → likely coordination or offset issues 

• High delay across all movements → possible cycle length or phasing issue 

This helps focus timing changes where they are likely to have the biggest impact. 

Step 5: Measure After Changes Are Made 

After retiming: 

• Recheck delay, arrivals on green, and excessive delay 

• Compare before/after periods using the same time windows 

• Confirm improvements are stable over multiple days 

This gives you defensible results to share internally. 

Step 6: Refine Your Changes 

One of the biggest advantages about probe-based signal performance measures is that they are always available. You can keep tweaking your timing plans constantly to find the sweet spot or adjust for minor changes over time. This gives you an opportunity for more iteration to find exactly what’s working without costly studies or in-field evaluations every time you change something.  

What “Good” Looks Like  

You’re usually looking for: 

• Fewer movements with excessive delay 

• Higher arrivals on green during coordinated periods 

• Reduced delay without creating new side-street problems 

• Consistent improvement across multiple days, not just one 

There’s no single “perfect” value, trends and repeatability matter most. 

Common Gotchas (Learned the Hard Way) 

• Construction weeks can completely distort results — exclude them through our API 

• Closely spaced signals need careful movement boundary definitions 

• Low volume movements may require longer time windows 

• Metrics show where issues are issues are, they don’t replace engineering judgment 

What to Hand Off  

Most cities export: 

• A one-page corridor summary 

• Before/after snapshots for key intersections 

• Notes explaining why certain locations were prioritized 

This makes it easier to align with consultants, management, or elected officials. 

Transportation agencies are expected to improve traffic operations despite limited staff, budgets, and field resources. Probe-Based Signal Analytics helps engineers quickly identify problem locations, prioritize retiming efforts, and measure results using continuous, objective performance data. 

By providing both network-wide visibility and intersection-level insights, it helps agencies spend less time finding problems and more time delivering measurable improvements for travelers. 

To learn more about Signal Analytics register for the INRIX upcoming webinar U.S. Signals Scorecard. 

FAQ  

Q: Do I need detectors or connected signals for this to work?
No. Probe‑based Signal Analytics relies on anonymous vehicle trajectory data and does not require roadside detection infrastructure. 

Q: Can I see my actual signal timing in the platform?
No. Signal timing data is not shown directly. Many agencies export analytics via API and combine them with their ATMS or signal system views. 

Q: Is this real-time signal optimization?
No. This is performance measurement, not adaptive control or optimization. 

Q: Can I estimate turning movements?
You can view turn ratios (left/through/right proportions). If results are close to warrants or thresholds, agencies typically confirm with an official count for liability purposes. 

Q: How does this compare to ATSPMs?
ATSPMs rely on roadside detection and timing data at equipped intersections. They also struggle to stitch intersections together to understand corridor performance. Probe‑based analytics measure how vehicles move and scale networkwide, even where detection doesn’t exist. 

Q: How often is the data updated?
Most agencies review results daily or weekly, using the same time‑of‑day windows for consistency.