Trafficure · Capabilities

Monitoring, alerts, analytics and reports, for the whole network

Trafficure turns Google Maps Roads Management Insights data into four working capabilities a city operations team uses every day: monitor every road, get alerted as congestion forms, read road-level analytics, and ship reports automatically. No cameras, no sensors, no construction.

Get a demo →

The shape of the platform

Four capabilities, one road graph

Trafficure is built on Google Maps Roads Management Insights, which delivers anonymized speed and movement signals aggregated per road segment from the more than one billion smartphones running Google Maps and Waze. RMI is the data layer. Trafficure is the activation layer that turns it into something an operations team can run a city on.

That activation layer resolves into four capabilities. Each one works off the same normalized road graph, so monitoring, alerting, analytics and reporting all speak about the same roads, the same segments and the same baselines. This page walks through what each does and what an agency does with it.

Capability 1

Monitor every road, not just the junctions with cameras

Trafficure monitors the whole network, arterials, residential streets and service lanes alike, normalized into more than 15,000 road segments per city against the road graph. The network refreshes on a two-minute cycle, so the operations picture is current rather than a snapshot from hours ago.

  1. 01

    Coverage · The full network is in view

    100% road coverage rather than the few hundred junctions a camera system can reach. The roads that never had a sensor are now monitored alongside the ones that did.

  2. 02

    Refresh · A two-minute cycle

    The whole network re-reads every two minutes, so what the command centre sees reflects the road now, not a stale reading.

  3. 03

    Resilience · No hardware to fail

    Probe data is unaffected by fog, rain, dust or darkness, and there are no cameras to decay. Coverage does not erode as equipment ages.

Capability 2

Be alerted as congestion forms, not after the complaints

Trafficure compares each road against its own baseline, a normal speed profile computed from roughly ninety days of history, and fires a predictive alert when a corridor begins to deviate. The alert lands while congestion is forming, which is the window in which an agency can still act on it. Alerts go out across the channels a team already uses.

Pre-emptive

Fires as congestion forms

Machine-learning alerts trigger on deviation from baseline, before a corridor is fully jammed, so a response is proactive rather than reactive.

Multi-channel

Reaches the team where they are

Command centre dashboard, WhatsApp and SMS, email, Slack, webhook and X. Thresholds, priority levels and auto-resolve are configurable per agency.

Closed loop

Surfaced and closed

An alert is not just raised. Build-ups are tracked through to clearance, so the operations team sees the event from formation to resolution.

Live congestion alerts in Trafficure: severe-congestion incidents listed by route with delay versus free-flow, and the selected alert highlighted on the road map
Live alerts: each incident scored against the corridor's own baseline.
Under 3 minalert latency, down from about 7, after the two-minute-cycle re-architecture
15–90 min → ~2 minincident-detection cycle, compressed from the citizen-complaint window

One road graph

Congestion, caught as it forms.

The same road graph that monitors the whole city is the one the alert reads from. A build-up is scored against that corridor's own baseline the moment it starts to deviate, and the alert reaches the team across every channel they already use.

Capability 3

Read the road at street level, with analytics you can act on

Trafficure normalizes more than 15,000 segments per city against the road graph and runs the full analytics cycle in under 120 seconds. The output is road-level, not a city-wide average: a CityPulse traffic-health score per corridor, a congestion ratio of free-flow speed to actual speed, and baselines built from roughly ninety days of history. That is the layer a planner uses to find a chronic bottleneck and prove an intervention worked.

  1. 01

    CityPulse · A traffic-health score per road

    A composite metric that turns raw speed into a readable health signal for each corridor, used directly in reports.

  2. 02

    Congestion ratio · Free-flow against actual

    The ratio of a road's free-flow speed to its current speed, so severity is comparable across very different roads.

  3. 03

    Baselines · Normal, learned from history

    Per-road baselines computed from roughly ninety days of history. Deviation from normal is what drives both the analytics and the alerts.

Trafficure · Corridor speed profile
Ring Road East · corridor C-12 · J01 → J07 eastbound
6.4 km · 7 junctions · working-day mean, 20 days to 08 Aug 2026 · probe fused with ATCC counts · run R-2026-2214
9.5 MIN LOST / TRIP
AM peak 17.9 min · free-flow 8.4 min · LOS E

Speed along chainage AM peak 08:00 to 10:00 vs free-flow

free-flow AM peak mean worst segments S3, S4
6040 200 km/h J01J02J03 J04J05J06J07 0.01.22.1 3.24.05.36.4 km 11 to 14 km/h

Segments eastbound · mean speed includes junction approach delay

SegLinkLengthFree-flowAM peakRatioTravel timeDelay
S1J01 → J021.2 km4834 km/h71%2.1 min0.6
S2J02 → J030.9 km4627 km/h59%2.0 min0.8
S3J03 → J041.1 km4614 km/h30%4.7 min3.3
S4J04 → J050.8 km4011 km/h28%4.4 min3.2
S5J05 → J061.3 km4829 km/h60%2.7 min1.1
S6J06 → J071.1 km4433 km/h75%2.0 min0.5
corridor6.4 km45.721.5 km/h47%17.9 min9.5 min

Junction diagnostics worst approach, AM peak

JnNameControlCyclev/cDelayLOS
J02Mill Gatesignal120 s0.8238 sD
J03Old Bridgesignal140 s0.9474 sE
J04Clock Towersignal180 s1.08178 sF
J05Depot Crosssignal160 s1.02151 sF
J06Lakeviewroundabout·0.7122 sC
Segments approaching J04 and J05 carry 6.5 of the 9.5 lost minutes (68%). J04 queue spills back 480 m into S3 by 08:40; J04 and J05 are 800 m apart and uncoordinated.

Corridor mean speed by hour peak spreading, quarter on quarter

300 23 km/h = 50% of free-flow 060912 151821 Q1 Q3
Below-50% windows, 15-min resolution: AM 07:50 to 10:10 this quarter against 08:20 to 09:40 in Q1, wider by 1h 00m; PM 16:50 to 19:10 against 17:10 to 18:40, wider by 50m.
Fig. · Corridor speed profile as rendered by Trafficure. All corridor, junction and speed data are illustrative and describe no real city, road network or customer.

Capability 4

Ship reports automatically, in standard formats

The analytics do not stop at a dashboard. Trafficure generates PDF reports built on the CityPulse scores and exports standard formats, so a weekly operations brief or a before-and-after intervention study is produced from the same live record rather than rebuilt by hand. Exports are standard formats, which means no vendor lock-in on the data an agency owns.

  1. 01

    Automated · PDF reports, generated

    Operations and planning reports are produced from live data, so the brief reflects the road graph as it stands rather than a hand-assembled snapshot.

  2. 02

    Portable · Standard-format exports

    Data exports in standard formats keep the agency's record portable, with no vendor lock-in on what the city owns.

  3. 03

    Evidence · Before and after, on record

    Because baselines persist, an intervention can be measured against the corridor's own history and the result written into a report.

Where the line is

What Trafficure is, and what it is not

These four capabilities are monitoring, alerting, analytics and reporting. Trafficure is not built for sub-second adaptive signal control, and it does not claim to be: that is where sensors and junction-level systems belong.

It runs on aggregated, anonymized movement with k-anonymity at the source, which means very sparse roads can drop out and there is no commercial-fleet-only or individual-level view. It is complementary to a camera ITMS, which keeps enforcement and number-plate recognition, while Trafficure covers the monitoring gap across the rest of the network.

Proof

Running in production, across cities

The four capabilities are not a roadmap. They run today in Pune, where Pune City Traffic Police became the first Indian city with software-only city-wide traffic management, and across the Kolkata region. In Pune the platform surfaced and closed 39,734 congestion build-ups with a median clearance of 26 minutes.

Surfaced and closed 39,734 congestion build-ups in Pune, with a median clearance of 26 minutes
  • Pune City Traffic Police
  • Pune Municipal Corporation
  • Kolkata Police
  • Howrah Police
  • Bidhannagar
  • Barrackpore
Get started

See all four capabilities on your own road network.

Bring your city's corridors. We will show you live monitoring, predictive alerts, analytics and reports running on Trafficure.

Get a demo →