Type at least two characters to search EyeSpot.

Search across solutions, detections, industries and resources ESC
AI VIDEO ANALYTICS FOR PUBLIC TRANSPORT & BUS FLEETS

Boarding. Alighting. Occupancy. Route intelligence.

AI video analytics for public transport and bus fleets, from onboard events to centralized fleet insight.

Existing camera infrastructure Flexible deployment options Defined and validated per scenario
Station platform monitoringDETECTION REPLAY
STN-12RECORDED
00:00 / 00:00
Visible detections00
Events to review00
Ready to play
Event reviewSTN-12

Loading recorded detections…

Rule: Passenger visibility · 0.75s confirmation

Recorded elevated view · Person detections and example review alerts.

Recorded demonstration with synchronized detections and example alert rules. Review the event details to see what this camera view can establish.

FLEET INTELLIGENCE

Every stop adds context.

Combine boarding and alighting events with route, stop, timestamp and device status. Explore the simulation to see how counts shape an occupancy estimate.

Doorway obstruction, unsafe crossing, supported fall detection and driver-monitoring modules require scenario-specific validation and configuration.

FLEET INTELLIGENCESIMULATED DATA
BUS 12 / ROUTE 151

Journey in progress

ESTIMATED OCCUPANCY27 passengers
BOARDING124
ALIGHTING97
Example StopExample location · Route active
Boarding +1 · Example Stop · 14:32:08SIM
Alighting +1 · Example Stop · 14:31:56SIM
Edge online2 cameras online
THE CHALLENGE

Start with the operational problem.

Fleet teams need a consistent view of passenger demand, doorway safety and device health across intermittent connectivity.

The EyeSpot approach

Process supported events onboard. Synchronize metadata and selected clips when connectivity returns, linking each event to bus, route, stop and time.

CAPABILITIES

What bus video analytics brings into view.

01

Boarding and alighting

02

Estimated occupancy

03

GPS correlation

04

Stop and route analytics

05

Doorway obstruction

06

Searchable incident clips

07

Device health

08

Remote configuration

09

Offline edge processing

10

Cloud synchronization

11

Crowd detection

12

Dwell-time analysis

13

Multi-bus management

14

Route history

15

Passenger flow

16

Unsafe crossing events

HOW IT WORKS

A clear path from capture to action.

Process supported events onboard. Synchronize metadata and selected clips when connectivity returns, linking each event to bus, route, stop and time.

01

Bus cameras

Doorway / cabin cameras

02

Edge AI device

Onboard edge device

03

Immediate detection

Immediate event detection

04

Events and clips

Metadata / selected clips

05

Cloud fleet platform

Fleet dashboard / APIs

Full raw video does not always need to be transmitted to the cloud. Edge processing can synchronize metadata, events and selected clips after connectivity returns.

YOUR DEPLOYMENT

Choose around your operating conditions.

INTEGRATIONS

Deliver context to the people who act.

Assess compatible camera streams and connect configured events to APIs, webhooks and notification workflows.

Integration options
VALIDATION & LIMITATIONS

Measure it where it will run.

Detection quality depends on camera placement, image quality, lighting and occlusion. Validate the configured scenario on representative footage before operational use.

Agree on review workflows, access controls and retention. Analytics support human decisions and do not independently establish safety or regulatory compliance.

See our evidence approach
COMMON QUESTIONS

Planning bus video analytics?

01How should we begin with bus video analytics?+

Process supported events onboard. Synchronize metadata and selected clips when connectivity returns, linking each event to bus, route, stop and time.

02Can we use our existing cameras?+

Compatible IP cameras and recorder streams can be assessed using their available protocols, codec, resolution and network access. Camera suitability is validated during the pilot.

03Does all video have to go to the cloud?+

No. Edge and on-premise processing can keep raw video local while forwarding metadata, events and selected clips according to your configuration.

04How do you validate a detection?+

We define the event, collect representative footage with permission and compare model outputs against manually reviewed examples. Lighting, occlusion and placement are included in the evaluation.

LET’S BUILD SOMETHING USEFUL

Your cameras already see it. Make it actionable.

Tell us about your environment. Let’s design the right video analytics pilot.

Book a live demo Your cameras. Your requirements. A clear next step.