RMIT University

Basic Information

Applicant Type: Organisation

Organisation Name: RMIT University

Main Questions

What is the problem you are solving and what is your solution?

Improper disposal of lithium-ion batteries in household waste has become a growing concern for municipal waste management systems. These batteries, commonly found in everyday electronics, pose a significant fire risk when crushed, punctured, or exposed to heat during garbage collection and compaction. When discarded in regular trash, they often end up in garbage trucks where mechanical stress and environmental conditions can trigger thermal runaway—a rapid, uncontrollable increase in temperature leading to combustion or explosion.

Fires in garbage collection trucks are particularly dangerous due to the enclosed and cluttered environment, which accelerates fire spread and hampers visibility. Detecting these fires early is challenging because traditional fire detection methods rely on visual cues or temperature thresholds that may not be effective in such conditions. The lack of direct line-of-sight and the presence of various materials can obscure early signs of combustion, delaying response and increasing the risk of damage, injury, and service disruption.

Early detection is critical to prevent escalation, but current systems often fail to identify fires until they are well underway. This highlights the urgent need for advanced, multi-modal detection technologies that can operate effectively in the complex environment of waste collection vehicles, enabling timely alerts and intervention to protect public safety and infrastructure.

Our solution integrates infrared (IR) and ultraviolet (UV) radiation detection with gaseous emissions monitoring, all connected through an Internet of Things (IoT) network and powered by artificial intelligence (AI). This multi-modal system is designed to operate before and during the initial stages of battery fire development, enabling early detection and rapid response.

- Gas sensors detect emissions unique to damaged lithium-ion batteries—such as those fractured or punctured—before visible combustion begins. These can include hydrogen fluoride (HF), carbon monoxide (CO), volatile organic compounds (VOCs), and phosphorus oxyfluoride (POF₃), which are not typically present in general household waste. In contrast, waste decomposition primarily emits methane (CH₄), carbon dioxide (CO₂), ammonia (NH₃), and hydrogen sulfide (H₂S). The presence of battery-specific gases, especially in the enclosed environment of garbage trucks, allows for more effective and targeted detection.

- IR and UV sensors monitor for increasing rates of change in radiation emissions, signaling the onset of combustion when there is a direct line-of-sight, particularly at the garbage entry.

- AI algorithms cross-reference data from all sensors to eliminate false positives and trigger timely alerts and automated response protocols.

By combining these complementary detection methods, our system ensures robust, reliable identification of battery fire risks at the earliest possible stage. This layered approach enhances safety, minimises damage, and supports proactive intervention in municipal waste collection operations.

Please outline the impact on Queensland and Queensland’s waste recycling industry - if the impact is in the community, please include it here. Important: this is not a community awareness raising challenge.

Implementing a multi-modal early detection system for lithium-ion battery fires in Queensland would significantly enhance the safety, efficiency, and sustainability of the state’s waste and recycling operations. Existing detection methods are generally reactive, activating only after a fire has started. They also struggle with false positives and limited visibility in the cluttered, enclosed environment of garbage trucks. By combining IR, UV, and gas sensors with AI and IoT, our solution extends to include pre-fire detection, multi-sensor cross-referencing, and real-time remote monitoring, which existing systems do not offer.

Industry Impact
Queensland’s waste sector faces increasing risks from improperly disposed lithium-ion batteries. These batteries, found in everyday items like phones, vapes, and power tools, are often discarded in household bins. When crushed or punctured during collection, they can ignite, causing fires in trucks and facilities. In 2024 alone, over 1,000 fires across Australia were attributed to lithium-ion batteries, with a significant portion occurring in waste and recycling environments.

Unlike traditional fire detection systems—which typically rely solely on temperature thresholds or smoke detection and often activate only after a fire is visibly underway—this solution introduces a layered, intelligent approach:
- Gas sensors detect emissions unique to damaged batteries (e.g., HF, CO, VOCs) before combustion begins.
- IR and UV sensors monitor for increasing radiation rates, signaling the onset of combustion.
-AI algorithms cross-reference sensor data to eliminate false positives and trigger timely alerts.

This multi-sensor, AI-enhanced architecture is a major innovation, enabling earlier, more accurate detection in the complex and enclosed environment of garbage trucks—where traditional systems often fail.

The solution aims to:
- Reduce Equipment Losses: Fires in garbage trucks often result in total vehicle loss and damage to surrounding infrastructure. Successful early detection would prevent such incidents, lowering insurance claims and repair costs.
- Improve Operational Continuity: Fire-related disruptions to waste collection routes and schedules would be minimised, ensuring consistent service delivery across Queensland’s urban and regional areas.
- Enhance Worker Safety: Waste collection and recycling staff are frequently exposed to fire risks. Early alerts would allow for safe evacuation and containment, reducing injuries and improving workplace safety standards.
-Protect Recycling Facilities: Damaged batteries reaching waste disposal facilities pose a major hazard. By detecting them earlier in the collection process, this system would reduce fire risks at these critical infrastructure points.

Community and Environmental Impact
The broader Queensland community would benefit from reduced fire incidents in residential areas, where garbage trucks operate daily. Fires in these settings pose risks to people, property, and traffic. Early detection would enable rapid response, preventing escalation and reducing the burden on emergency services.

Economically, the solution would contribute to the resilience of local councils and waste contractors. By reducing fire-related losses and improving system reliability, it would stabilise service costs and allow better allocation of public funds.

The risk of fire from improper battery disposal grows. Adopting this innovative detection system would position Queensland as a leader in waste safety innovation, protecting infrastructure, workers, and communities, while supporting a more sustainable wastey industry.

Please outline details about your organisation and your proposed business model with supporting documentation, include how many units need to sell to launch as a business

Organisation Overview
Name: Royal Melbourne Institute of Technology (RMIT University)
Legal Structure: Public Research University
Location: Melbourne, Victoria, Australia
Founded: 1887
Campuses: Melbourne City, Brunswick, Bundoora (Australia); Ho Chi Minh City, Hanoi (Vietnam); Barcelona (Spain)
Website: rmit.edu.au
Mission and vision: RMIT University’s research and industry engagement mission is to drive impactful, real-world innovation through collaborative partnerships and applied research. Guided by its strategic vision, Knowledge with Action, RMIT focuses on solving global challenges in areas like emerging technologies, smart cities, and social innovation. The university integrates industry into education and research, fostering co-creation with businesses, communities, and global partners to ensure its work is practical, scalable, and future-focused.

RMIT University in collaboration with SumoEcoTech, an environmental technology company focused on developing innovative solutions for waste management and sustainability, is pioneering a multi-modal fire detection system for municipal garbage disposal trucks. This partnership combines SumoEcoTech’s industry expertise in sensor technology with RMIT’s research capabilities in artificial intelligence and fire testing.

Proposed Solution
The proposed solution addresses the growing risk of lithium-ion battery fires in municipal waste collection vehicles. It integrates infrared (IR) and ultraviolet (UV) radiation sensors with gas detection modules to identify pre- and early-stage fire indicators, including emissions unique to damaged batteries. These sensors are connected via an IoT network and analysed using AI algorithms to cross-reference data, eliminate false positives, and trigger timely alerts. This layered approach enables detection before visible combustion, significantly improving safety and operational continuity – A capability not offered by current market solutions.

Business Model
The business model is structured around hardware sales, software licensing, and ongoing support services. Revenue streams include:
- Sensor kit sales per vehicle
- AI analytics platform subscription
- Installation and integration services
- Maintenance and calibration contracts
- Data analytics services for fleet-wide fire risk insights

Target customers include municipal councils, private waste contractors, and fleet management providers. Sales channels will include direct sales, strategic partnerships, and industry events. Customer relationships will be supported through dedicated account management, training, and technical support.

Launch Requirements
To launch the business, we estimate the need to sell and deploy approximately 100 units. This volume covers initial manufacturing costs, software development, and operational setup. Each unit includes a sensor kit, IoT module, and access to the AI platform. This initial deployment will also serve as a pilot program to validate system performance in Australia and generate case studies for broader market adoption.

Market Readiness

As batteries becoming more utilised in modern engineered products and applications, their disposal is increasingly problematic. Correct disposal campaigns and establishment of dedicated disposal locations have proven to be of limited effectiveness. Consequently, the associated risk of battery fires in the waste disposal stream is significant. Fires have been reported in every aspect of waste disposal from bins and disposal vehicles through to treatment facilities and landfills. These fires pose a real risk to the safety of the public, operators of waste disposal equipment/facilities, and to the equipment and effective functionality of waste disposal facilities.

Globally, there is no consensus in the waste disposal sector on how to effectively manage the presence of batteries in the waste disposal stream. Many solutions rely on pre-screening via optical or X-ray scanners before the waste is processed in machines such as shredders/balers which pose a heightened risk of causing battery damage and the initiation of fire. Although these methods can be effective at protecting the waste disposal facility and their operators, there is no impact on the vulnerability of waste disposal collection vehicles and drivers to battery fires.

There is a growing body of research for the identification and quick response to battery fires in waste collection vehicles. Implementation of high TRL technology such as scanning (e.g., X-ray computed tomography), are not feasible in the immediate future due to the high financial cost. Therefore, the early detection and response to battery fires is critical to protecting the drivers and their vehicles. However, the compact and confined nature of the vehicle makes identification of the fire event extremely difficult. Traditional scanning methods (e.g., optical, infrared cameras) are of TRL up to 9 and can be readily implemented. However, these are less effective than when used in the waste disposal facility because often there is no direct line of sight to the ignition site. Consequently, detection and response cannot occur before or in early-stage fire development and the response of the driver and emergency services is going to be to an established fire when risk is highest, and extinguishment is hardest. This project focuses on the development of technologies to enable response during the pre- or early-fire stages to address the battery fire risks at the point of collection to increase protection along the entire waste management line.

The technology currently sits at a TRL of 6 with technology demonstrations having been completed in Korea by partner organisation SumoEcoTech. This project expects to provide a technology demonstration in the Australian context and provide further technology development, particularly in the AI and multi-modal functionality. The expected TRL at the completion of the project is 8.

To prepare for commercial readiness in Australia, the following milestones need to be completed:

Milestone 1: Verification of compliance and compatibility in the Australian context.
In this milestone, the hardware components will be verified for compliance to Australian Standards and the compatibility of the equipment to be installed with Australian waste disposal trucks will be assessed. The outcomes of this milestone include a detailed report detailing the design requirements required to install the multi-modal detection system in Milestone 2.

Milestone 2: Design of integration and installation method for Australian trucks.
The second milestone involves the investigation into the AI and IoT integration to enable the response output required by the Queensland waste disposal sector. The installation method specific to the waste disposal trucks used in Queensland will be developed and any design adjustments for compliance identified in Milestone 1 will be implemented.

Milestone 3: Installation and in-situ functionality and pre-commercial demonstration.
The installation of the multi-modal detection system on an actual waste disposal vehicle will be the focus of milestone 3. It continues the design and testing started in milestone 2 with a focus on the in-situ performance and system response functionality. Once established, a pre-commercial demonstration of the multi-modal detection system will be performed for stakeholders.

Milestone 4: Pre-commercial activities and product launch.
This milestone is centred around pre-commercialisation activities including risk assessment of the equipment, evaluation of operational costs, and dissemination of the detection capabilities to the industry. The aim of this milestone is to prepare for the implementation in Queensland and future extension to other Australian states and territories as the final step of commercialisation.

The estimated timeframe for product development are as follows:
- Estimated number of weeks notice needed to run a pilot: 52 weeks.
- Estimated number of weeks the pilot would take: 12 weeks.
- Estimated number of weeks from pilot to commercial launch (ie: how long would it take to run a pilot and how long before ready to sell product to market?): 26 weeks.

Please provide evidence of your team’s experience of commercial experience and project delivery

Dr. Thomas Loh: Loh has established a strongly industry focused research career (3 CRC-P, 1 Innovation Connections). He has also spent a significant portion of his career providing direct consultation to industry (6 projects). Combined ~50% of his career has focused on industry-led projects. He has also dedicated a significant proportion (~30%) of his career to working with Defence to implement engineering solutions to defence platforms; early in his career on an SBIRD project with DSTG and more recently employed directly with DSTG. He has important experience in research translation having contributed significantly to the commercialisation of ROBOVOID and ongoing co-leadership in establishment of Australia’s first pilot plant to recycle composite wind turbine blades with industry. Loh will contribute to all project phases with his interdisciplinary and industry-oriented expertise in applied engineering and research translation.

Prof. Kate Nguyen: Nguyen specialises in artificial intellengence and fire safety engineering, focusing on developing innovative fire-retardant materials derived from industrial waste. She has extensive experience in leading major projects, including over $9 million in industry-collaborated research funding, contributing to 2 pilot plants and 3 patents. Nguyen has been recognised as Australia’s No#1 researcher in composite materials (2021, The Australian Research Magazine), one of ten female researchers leading the fight against climate change (2022, Global Citizen) and Victoria Young Tall Poppy Award (AIPS, 2023) for her effort in bridging academia and industry. Apart from her role in academia, she was appointed in 2021-2024 as the Chief Engineer at Cladding Safety Victoria, Victoria State Government, where she led the development of the world's first risk-based cladding mitigation protocols, providing practical remedial solutions for over 300 buildings. She is currently a Chartered Engineer and Australia’s representative on the Steering Committee of the United Nations’ International Fire Safety Standards Coalition. Nguyen will contribute to all project phases with her interdisciplinary and industry-oriented expertise in fire engineering, materials science, and large-scale testing.