Roads
Disconnected signal + camera networks
Congestion data sits in one tool · cameras in another · re-routing happens manually after delays.
- Cameras siloed from CAD
- Manual diversion plans
- Late incident impact
// Civic surfaces · Smart Mobility
Optimize traffic, transit, parking, EV, and autonomous infrastructure.
A connected mobility command layer for roads, transit, parking, EV chargers and future autonomous corridors. Reduce congestion, improve service reliability, increase infrastructure uptime — with one real-time operating view.
5
Mobility surfaces
AI
Re-route advisor
Real-time
Signal data
TrafficTransitParkingEV
// 01 · What operations look like today
Today: stitched systems, slow response.
Transit
Passenger experience guesswork
No live picture of station crowding, route delays or passenger flow — riders get stuck, operators are blind.
- Station crowding unknown
- Delay alerts via radio
- Inter-modal hand-off lost
EV
Charger uptime unmanaged
Field teams discover faults from complaints — not from sensor data.
- No fault telemetry
- Reactive maintenance
- Wait-time invisible
// 02 · What changes with Innfini
With Innfini: live, AI-assisted mobility ops.
Traffic
Live traffic intelligence
Real-time heatmaps, incident impact, signal coordination + AI re-route advisor with cited reasoning.
- Heat & flow lines
- Closure workflows
- AI re-route
Transit
Transit coordination
Bus, metro, tram or shuttle status, route delays, station crowding and passenger flow on one canvas.
- Service interruption alerts
- Field dispatch
- Passenger flow
EV & AV
Charger + AV readiness
Live charger health, demand hotspots, queue time. Roadside units and AV corridors mapped + monitored.
- Faults · utilization
- Demand hotspots
- AV-corridor sensors
// 03 · The screens
Eight surfaces, one operating layer.
Each Innfini screen below is an illustrative scenario from one downtown grid — the same map, the same event model and the same audit trail from traffic to EV to autonomous corridors.
01 · Traffic intelligence
Congestion is detected, a diversion is modelled, and an operator decides.
Radar and camera analytics watch every signalised junction. When a queue forms at 5th & Park, Innfini scores the corridor, models an alternative via 7th Avenue that avoids the school zone, and puts the proposal — with the radar and camera evidence beside it — in front of the operator before it touches a single signal.
- Live congestion from loops, radar, ANPR and CCTV analytics
- Diversion proposals modelled and policy-checked · school zones avoided
- Signal coordination across the corridor · junction by junction
- Evidence attached · radar trace and camera view for the decision
02 · Transit coordination
Metro, bus and tram on one interchange view — with the transfer walk in between.
AVL, passenger counts and GTFS-Realtime feed a single picture of Central Station: which services are running, which connection is under review, where a tram delay lands, and how full the plaza is right now. Operators hold, short-turn or add capacity and riders are told at the same moment.
- Service status per mode · running, connection review, delay reported
- Station occupancy heatmap · platforms, plaza and transfer walks
- Route status Central → Riverside → East Park
- GTFS / GTFS-RT, SIRI and AVL in and out · rider apps stay current
03 · Parking & kerbside
Spaces, loading bays and guidance — published from the same source.
Bay sensors, barrier counts and kerb permits are fused into one availability picture. A driver is guided to a free garage level while a loading bay is held for the van that booked it; signage and city apps are updated from the same event, so what the street shows is what the system knows.
- Garage availability by level · available, occupied, reserved
- Kerbside loading zones · permits and time windows enforced
- Guidance published to signage and apps automatically
- CDS kerb data specification · payment and enforcement integrated
04 · EV charging
A charger fault becomes a maintenance job — and the availability map is already updated.
OCPP telemetry from every charge point lands on the map. EV-07 reports a fault; Innfini confirms the telemetry, assigns maintenance with the next steps, and updates the published availability so drivers are not sent to a dead charger. Site demand is tracked so load can be rebalanced across the estate.
- OCPP 1.6J / 2.0.1 telemetry · charging, available, service required
- Fault review with next steps · diagnosis, module, return to service
- Availability published to drivers the moment status changes
- Site demand curve · rebalance load across multi-CPO estates
05 · Corridor readiness
Roadside units, sensor zones and geofences — certified before the first shuttle runs.
An autonomous corridor is a live digital twin: every roadside unit, its sensor zone and the geofence it belongs to, checked continuously. RSU-07 drops into maintenance and the readiness state moves to "under review" until it is back — so the shuttle only runs on a corridor that is proven ready.
- Roadside units connected · V2I secure link to each vehicle
- Sensor zones monitored · lidar, camera and signal-phase data
- Geofence defined per corridor · pilot to citywide
- Readiness state with maintenance flags · disengagements tracked
06 · Incident response
Detect & confirm → assess & propose → approve & coordinate.
A collision at 5th & Park is flagged by camera and sensor together. Innfini proposes the diversion via 7th Avenue with the impacted lanes closed, then waits: a named operator approves, dispatch and notifications go out, and every step — with the camera frame and radar trace — lands on the event timeline.
- Camera + sensor confirmation · no operator has to go looking
- Diversion proposed with lanes closed and transit informed
- Operator approval required before anything acts on the street
- Evidence and events · camera, radar trace, timestamped timeline
07 · Connected mobility
Traffic, transit and EV & kerb on one shared event model.
Signal control, traffic flow, next arrivals, station status, charging usage and kerb management all flow into the same event model — so a signal change, a bus delay and a charger fault are the same kind of object, correlated on the same map, and exposed through the same APIs to the systems the city already runs.
- NTCIP signal controllers · GTFS-RT transit · OCPP and CDS for EV & kerb
- One shared event model across sensors, vehicles, infrastructure, APIs
- Cross-surface correlation · an incident shows up in transit, too
- Open data published back · no rip-and-replace
08 · Operator control
Nothing acts on the street without a policy behind it and a name against it.
A lane closure on Riverdale Bridge produces a proposed diversion around the Riverview School protected zone. Before an operator can approve it, Innfini shows that the evidence is linked, the policy check passed and a field team is assigned — and the audit chronology records who created, checked and approved it.
- Diversion review · pending approval until a named operator acts
- Policy guardrails · protected zones and restricted hours respected
- Evidence linked · camera, incident and traffic conditions
- Audit chronology · proposal, policy check, approval, immutable
// 04 · The five mobility surfaces
One layer. Five surfaces.
Each surface runs on the same event model, the same map and the same audit trail — so a signal change, a bus delay and a charger fault are the same kind of object.
| Surface | What it ingests | What operators do | Typical scale |
|---|---|---|---|
| Roads & signals | Loop counters, radar, ANPR, adaptive controllers, CCTV analytics | Retime corridors, push diversions, close lanes with a workflow | 400–2,000 intersections |
| Transit | AVL/GPS, APC passenger counts, GTFS-RT, depot telematics | Hold, short-turn, add capacity, notify riders | 10–400 vehicles per mode |
| Parking | Bay sensors, barrier counts, payment systems, kerb permits | Dynamic pricing, guidance signage, enforcement tasking | 3k–60k stalls |
| EV charging | OCPP 1.6/2.0.1 telemetry, fault codes, session records | Dispatch maintenance, rebalance load, publish availability | 50–5,000 points |
| Autonomous corridors | Roadside units, V2X, lidar zones, HD-map deltas | Certify corridors, geofence, monitor disengagements | Pilot → citywide |
// 05 · Incident to clearance
What happens in the first four minutes.
A collision on 5th & Park, from detection to cleared lane — every step time-stamped and attributable.
00:00 → 00:12
Detect & confirm
Radar flags a speed collapse; the nearest camera confirms it on a vision score, so no operator has to go looking.
- Speed drop > 60% in one cycle
- Camera confirmation 0.91
- Incident opened automatically
00:12 → 00:50
Assess & propose
The platform scores severity from lane blockage, queue growth and downstream demand, then drafts a diversion.
- 2 lanes blocked · queue +180 m/min
- Diversion 5th → 7th modelled
- −18% network ETA impact
00:50 → 04:00
Act & clear
Operator approves once. Signals retime, signage updates, EMS is dispatched and transit is told to hold.
- Corridor retimed · 6 junctions
- EMS-04 on scene 4:08
- Riders notified in-app
// 06 · Standards & integration
It speaks what your city already runs.
Traffic
Controllers & detection
NTCIP 1202/1203 signal controllers, TMDD centre-to-centre, SPaT/MAP broadcast, Datex II feeds.
- Adaptive vendor agnostic
- Existing loops and radar reused
- No rip-and-replace
Transit
Service data
GTFS and GTFS-Realtime in and out, SIRI feeds, depot and AVL telematics, fare-system events.
- Rider apps stay current
- Open data published back
- Multi-operator supported
EV & kerb
Charging & kerbside
OCPP 1.6J and 2.0.1, OCPI roaming, CDS kerb data specification for permits and loading zones.
- Multi-CPO estates
- Fault codes normalised
- Availability published
// 07 · Governance
Every automated action is attributable.
Traffic control is a public-safety function. Nothing acts on the street without a policy behind it and a name against it.
| Control | How it works |
|---|---|
| Human-in-the-loop by default | AI proposes; a named operator approves. Auto-execution is opt-in per action type and per corridor. |
| Policy guardrails | Diversions cannot route heavy vehicles onto residential streets or past a school in restricted hours. |
| Full audit trail | Every proposal, approval, rejection and signal change is written immutably with the evidence that triggered it. |
| Explainability | Each recommendation cites the sensors, thresholds and prior incidents it was derived from. |
| Privacy | ANPR and video are processed for counts and classification; identity data is not retained by the mobility layer. |
// Why it matters
Measured outcomes.
−24%
Congestion response
From incident detection to diversion live.
+12%
Transit reliability
On-time ride share after signal AI rollout.
+18%
Charger uptime
Predictive maintenance + live fault telemetry.
AV-ready
Future-ready
Roadside units · sensor zones · AV corridors.
// In summary
Innfini Smart Mobility gives cities a connected command layer to monitor, optimize and coordinate movement across roads, transit, parking, EV and autonomous mobility.
See it on your city.
A walkthrough with our solutions team — your data, your departments, your workflows.