XTrack Vision AI
A camera that understands what's happening at the vehicle
Vision AI turns footage from onboard cameras into telematics events: collections, violations, asset status. No additional sensors, no footage review.
VisionAI evolves classic video monitoring toward an intelligent sensor. The camera is no longer used solely to record footage — combined with XTrack’s proprietary algorithms, it automatically recognises events, people, and assets, and interprets operational situations, turning footage into data that can be used directly in day-to-day waste collection management.
The solution addresses one of the industry’s key challenges: a scale of operations that can’t be effectively controlled manually. A single vehicle can perform hundreds, even over a thousand bin operations per shift. VisionAI automatically singles out the events among them that require attention, documentation, or a response.
See it in action
XTrack Vision AI on board a refuse collection vehicle
See how the onboard camera recognises operations, detects violations, and feeds the system with data — in real time.
RCV
9 XTrack VisionAI functions for refuse collection vehicles
Regulations
Automatic facial anonymisation in camera recordings
XTrack VisionAI automatically anonymises the faces of people appearing in onboard camera recordings — both during live viewing in the system and in exported material.
- Automatic face detection and anonymisation
- Real-time anonymisation during playback in XTrack WebX
- Anonymization in material downloaded by users
- Precise access management: recordings with anonymisation or in the original version
Safety
Detecting crew on steps and body elements while driving
The system recognises people in hazardous zones of the vehicle and logs the event together with the footage.
- Automatic health & safety rule enforcement
- Instant notification of a violation
- Material linked to the vehicle, time, and location
Collection
Manual waste loading into the RCV bypassing control modules
The camera detects manual loading and flags operations performed outside the standard collection process.
- Control of non-standard collections
- Fast complaint verification
- Event available without reviewing the entire recording
Collection
Bin type recognition (as a supplement to, or replacement for, the RFID module)
The algorithm recognises the type of collected bin and links the result to the specific collection operation.
- Identification without an additional sensor
- Supplements RFID data
- Consistent collection history in XTrack
Collection
Detecting bin attachment to the lifter, without additional sensors
Vision AI confirms bin attachment to the lifter based on the camera image.
- Confirmation of operation completion
- Fewer onboard infrastructure components
- Data ready for billing and reporting
Collection
Detecting fraction contamination
The system flags visible contamination of the collected fraction and creates event documentation.
- Automatic detection of deviations
- Documentation for complaints
- Better control of raw material quality
Collection
Fraction compliance control
Image analysis helps confirm whether the collected material matches the planned fraction.
- Early detection of mistakes
- Linking footage to the route and schedule
- Faster resolution of discrepancies
Collection
Detecting overfilled bins
The camera recognises overfilling and logs the event before the collection operation begins.
- Objective record of bin condition
- Data for schedule optimisation
- Fast handling of resident reports
Diagnostics
Automatic camera cleanliness control
Vision AI detects reduced camera visibility before image quality affects event analysis.
- Early notification of contamination
- Fewer gaps in documentation
- Continuous image quality control
Container trucks
4 Vision AI functions for hooklift and skip-loader operators
A container is the only asset in this industry that a company knowingly lets out of its hands for several days. Vision AI turns the number on the body and the body’s activity into asset status.
Operations
Container location based on XTrack Vision ID number reading via the camera
The camera reads the container number off its side and links it to the operation — no GPS locator on the container, and no manually copying numbers onto a waybill.
- Up-to-date location of every container, including one that’s been sitting at a client’s site for a week
- Alert on an idle, non-rotating asset instead of buying equipment that seems to be “missing”
- Documented issue and return per number as a basis for enforcing liability for loss
- Charging demurrage fees based on actual drop-off and pick-up dates
Operations
Detecting container loading and unloading
The system recognises the direction of the operation — drop-off or pick-up — and knows whether something is currently on the body.
- Real-time body status: the dispatcher directs the vehicle that actually has a free frame
- Detecting runs and containers outside the work order
- Measured share of empty runs as a basis for reducing them
- “While you’re at it” orders: containers in the field matched with vehicles already running empty
Regulations
Automatic image anonymisation
From a single recording you get, in parallel, both the full material and the anonymised one — no frame-by-frame editing.
- Meets the anonymisation requirement set out in the tender specification, without extra tools or staff time
- The evidentiary material stays unedited; you issue the anonymised version externally
- Permissions determine who sees which version of the recording
- Recording release measured in hours, not days — even with four streams from the vehicle
Diagnostics
Camera cleanliness control
The system itself reports lens contamination or obstruction before the day’s footage loses its evidentiary value.
- Clean the camera and continue the route instead of returning to base
- Continuity of recording and operational data throughout the workday
- You learn about the problem the same day, not at the complaint stage
XTrack Vision AI
One platform — data + video
No more searching through hundreds of hours of footage. Every event detected by Vision AI becomes an integral part of the vehicle’s telematics data — you filter it, sort it, export it, and build alerts on it exactly as you would with core GPS data. The system automatically links the selected video clip from the XTrack Cloud, and makes everything additionally available via the XTrack REST API to your billing/ERP systems.
XTrack Vision AI
Benefits for the user
The common denominator of all XTrack VisionAI functions is automating tasks that traditionally require human work or additional devices and sensors. The key benefits are:
Automating operation control and reducing manual video footage review
Greater crew safety through automatic detection of defined hazardous situations
Lower risk of losing assets and more efficient use of the container fleet
Proof of service completion linked with the data for the specific event
Better control of the collection process and faster detection of irregularities
Logging container loading and unloading, and building a usage history
Simplified onboard infrastructure — a single camera with AI algorithms can perform some of the functions that previously required additional devices and sensors
Case studies
Cases, law, and industry practice
See how every XTrack Vision AI function looks from the perspective of real legal requirements, tender specification provisions, and situations that occur on the route every day.
The new safety standard in waste collection
ZÖLLER-KIPPER GmbH, a European leader in the production of municipal vehicles, integrates XTrack VisionAI technology into its Z-CAM system, which autonomously protects operators in the lifter zone.
Want to see how our system works?
We’ll show you our platform in practice, using examples relevant to your industry, processes, and field resources.
Contact