VisionHQ Pty Ltd

Basic Information

Applicant Type: Organisation

Organisation Name: VisionHQ Pty Ltd

Main Questions

Problem Solution

Waste collection trucks face significant risks when hazardous materials—such as car batteries, LPG bottles, and gas canisters—are incorrectly disposed of into the hopper. These materials can:
-Ignite or explode during compaction
-Cause severe damage to high-value trucks and onboard equipment
-Create serious safety hazards for drivers and nearby personnel
-Lead to costly downtime, service disruptions, and rising insurance premiums

Currently, the industry relies primarily on fire or thermal detection systems that activate only after a hazardous event has already begun—too late to prevent damage or harm.
There is no automated, real-time detection system widely deployed on waste trucks that can identify hazardous items at the point of disposal, when early action is still possible.

The risks extend downstream as well. If dangerous goods pass through the collection stage undetected and reach recycling or recovery facilities, they can:
-Cause fires in shredders or processing lines
-Damage expensive stationary equipment
-Endanger workers on-site
-Trigger shutdowns and cleanup costs

The waste and recycling industry in Queensland operates as a complex chain—from kerbside pickup to material recovery—yet this system remains vulnerable to instability caused by undeclared or dangerous items. With limited resources and few fire suppression tools available onboard trucks, councils and waste contractors lack the ability to manage these risks at the earliest, most effective point: the truck hopper.

This gap in early detection presents both a critical operational need and a significant market opportunity for AI-driven, real-time hazard detection on waste collection vehicles.

Impact

VisionHQ has successfully developed and commercialised a contamination detection system for waste collection vehicles, capable of identifying incorrect waste in both green and recycling streams. This system uses a camera positioned inside the hopper, which can either be newly installed or integrated with the truck’s existing CCTV system. It operates entirely on an isolated edge computing device with minimal infrastructure and data transmission requirements making it highly scalable across fleets.

To address the more complex challenge of hazardous waste detection, this system will be enhanced with improved imaging sensors (not just cameras) and faster edge processors.

The upgraded Hazardous Object Detection System will feature:
-Real-time identification of dangerous waste items during collection
-Immediate alerts to drivers before compaction (driver-assist notification only, no automated intervention)
-Communication-ready design to potentially interface with suppression systems in the future
-Automated evidence logging for compliance and safety audits

This innovation closes a critical safety and compliance gap in Queensland’s waste and recycling sector. It offers a practical way to prevent fires, equipment damage, and workplace injuries—issues that are costly and increasingly frequent due to undetected dangerous items.

By detecting hazards at the earliest point in the waste chain—the truck hopper—The system protects downstream processes and reduces operational burden at material recovery facilities. It is also equally suitable for use within these facilities, catching what may have been missed during collection.

VisionHQ’s solution is hands-off for drivers, acting as an assistive tool that enhances awareness without requiring intervention, ensuring compliance without adding operational complexity.
*Impact on Queensland and the Recycling Industry
-Operational Safety
Functioning as a driver-assistance hazard detection tool, the system helps drivers identify dangerous items in real time during collection. Early alerts reduce the risk of fires, explosions, and machinery damage, protecting workers, vehicles, and infrastructure alike.
-Scalability
The AI-based solution is compact, infrastructure-light, and easily deployable across council fleets, enabling rapid adoption in both urban and regional communities without significant operational disruption.
-Cost Reduction
By automating hazard detection, the system reduces manual inspection time, incident-related downtime, and insurance liabilities—delivering measurable savings to councils and waste contractors.
-Data-Driven Compliance
All detections are recorded with timestamped image evidence, supporting regulatory reporting, environmental compliance, and incident traceability—without interfering with existing waste collection workflows.

*Community Benefit
While not a community engagement or awareness tool, the system delivers clear indirect benefits to the broader community through safer, more reliable waste services.

By automating hazard detection at the truck hopper, the system reduces the risk of fires, explosions, and operational incidents in residential areas—protecting both the public and frontline waste workers. It also removes the need for drivers to manually monitor camera feeds, reducing distraction and improving focus during collection.

In turn, this leads to:
-Fewer service disruptions in neighbourhoods
-Lower risk of injury or damage in densely populated areas
-Improved environmental outcomes through early removal of hazardous materials from the waste stream

Overall, the system supports more responsible, efficient, and community-safe waste operations across Queensland.

Business Model

Organisation Overview
VisionHQ is a Brisbane-based AI technology company specialising in edge-based computer vision solutions for the waste collection industry. Our core mission is to automate compliance, safety, and operational insights using smart camera systems mounted on waste collection vehicles. VisionHQ is currently deployed across multiple Queensland councils, with proven success in AI-powered contamination detection and street-level infrastructure monitoring.

Proposed Business Model
VisionHQ operates on a Hardware-as-a-Service (HaaS) model. Councils and waste contractors pay a monthly subscription per truck, covering hardware rental, real-time AI processing, secure cloud access, and full technical support. The system is plug-and-play, scalable, and designed to require minimal IT overhead from the customer.

Typical Pricing Model
-Hardware Rental: AUD $79/month for 48 months
-AI Detection Software License: AUD $299/month
-Total per truck: AUD $378/month

Go-to-Market Plan
We will prioritise deployment in high-risk collection areas where hazardous waste incidents are most likely to occur. VisionHQ plans to commercialise the hazardous object detection system through strategic partnerships with councils and waste contractors, targeting an initial rollout of 30 trucks supported by internal capital and early pilot revenues.

Break-even Point
The business reaches break-even at approximately 100 trucks, generating AUD $37,800 in monthly recurring revenue—sufficient to cover operational costs including engineering, customer support, cloud infrastructure, and hardware servicing.

Market Readiness

The hazardous object detection system builds on VisionHQ’s existing and proven contamination detection platform, which is already commercially deployed across several councils in Australia. The core detection engine, data pipeline, and edge deployment infrastructure are functionally complete and actively in use.

Current Deployment Status (Contamination Detection):
-City of Belmont (WA) – 5 trucks (paying)
-Logan City Council (QLD) – 13 trucks (paying)
-Brisbane City Council (QLD) – 5 trucks (trial)
-Cleanaway Melbourne (VIC) – 1 truck (trial)
-Solo Resource Recovery (NSW & SA) – 2 trucks (trial)

To extend the system for hazardous waste detection, VisionHQ will undertake a structured upgrade involving higher-resolution cameras, faster edge compute units, and the collection of a new AI training dataset. While the underlying platform is mature, these enhancements are required to reliably identify more complex waste items like batteries and LPG bottles.

Roadmap:
01/06/25 – Deployment of upgraded camera onto pilot truck for initial AI training data collection
01/07/25 – Development and training of high-resolution AI model
01/08/25 – Deployment onto Brisbane City Council fleet (28 trucks) for contamination detection
Q1 2026 – Collection of past hazardous waste events for training specialised AI model
Q2 2026 – System testing with mock hazardous waste items
Ongoing – Refining and polishing of solution

While the existing platform can be considered Technology Readiness Level (TRL) 9, the inclusion of upgraded sensing and model architecture for hazardous material detection brings the readiness of the full solution to approximately TRL 6–7, pending further pilot validation and data maturity.

Team

VisionHQ brings together a multidisciplinary team with deep expertise in AI, edge computing, cloud infrastructure, and waste industry applications. The team combines practical engineering capability with real-world deployment experience across Australian councils and waste operators.

*Robin Lin – Founder & Director
13+ years of experience in the mobile CCTV and waste tech industry
Founder of Fleetway (2015), a leading platform for fast CCTV data retrieval from waste trucks
Master of Digital Design (web screen)
Bachelor of Information Technology

*Allen Liang – Technical Lead
6+ years of experience in the ICT industry
Certified Google Cloud Professional Developer
Cisco Certified Network Associate
Master of Computer Science
Bachelor of IT and Bachelor of Business (Marketing)

*Alex Smith – Full Stack Developer
5+ years of experience in front-end and back-end software development
Skilled in building scalable web applications and managing AWS infrastructure
Experienced in security and performance optimisation using Cloudflare
Bachelor of IT (Computer Science)

*Daniel Zhang – AI Engineer
3+ years of combined academic and industry experience
Specialised in MLOps, DevOps, and edge AI deployment
Experienced in configuring AI workloads on Google Cloud and embedded devices
Bachelor of IT (Computer Science)

*Duval Longa – AI Engineer
3+ years of experience in applied AI and cloud-based ML pipelines
Co-author of peer-reviewed ML research in material property analysis
Skilled in MLOps, data engineering, and AI performance optimisation
Bachelor of Engineering (Software)