Posts by Michelle Luca

Securing AI interactions: enhancing the MCP with governance, authentication, and real-time monitoring

This project strengthens the model context protocol (MCP), a framework designed to enforce contextual integrity in AI-driven interactions throughout their lifecycle, to improve AI security by addressing gaps in governance, authentication, and monitoring. The outcome will be a robust security framework for MCP, enhancing AI system trustworthiness. This solution will improve AI security and reliability, benefiting industries and enabling safer AI deployment.

Investigating the potential applications of biomass and waste gasification by-products

This project will investigate challenges and potentials of biomass and waste gasification by-products such as ash, slag, and biochar for alternative applications. There are several potential applications for these materials in different industries, including cement and asphalt. The expected outcomes are filling the gap in the knowledge on the technical challenges of using the biomass and waste gasification by-products. This project will also provide a thorough understanding of the commercial feasibility of the proposed applications for these by-products. Potential benefits include finding techno-economically feasible applications for utilising gasification by-products that will make the overall process zero-waste. This will also help the off-taking industries in their decarbonisation efforts by using more sustainable resources. 

Exploring autonomous methods for distribution network identification

This project aims to develop advanced AI-driven methods to automate distribution network identification, addressing challenges from the growth of zero-emission energy resources. The expected outcomes include innovative study methods and prototype software. Potential benefits span enhanced network management and seamless resource integration, accelerating Australia’s shift to a carbon-neutral energy future.

Development of a novel system for daily patient dose verification using a transmission detector

This project will develop a novel system that uses a transmission detector to provide accurate and independent verification of the dose delivered to cancer patients receiving radiotherapy. The expected outcomes are to produce an estimation-based algorithm to determine the treatment dose, reconstructed on daily verification images and a sensitivity analysis by introducing known delivery errors and validating results against independent measurements. This technology will enhance the safety and accuracy of modern radiotherapy, enabling personalised treatment assessment and improving patient outcomes

Exploring solar-powered EV charging networks

This project explores strategies for developing a solar-powered EV charging network that ensures comprehensive coverage across urban and remote regions in Australia. The expected outcome is to enhance energy security, mitigate range anxiety, and protect user data, by leveraging AI-driven data analytics and privacy-preserving mechanisms. The project will potentially provide insights into the feasibility and design of a scalable and sustainable EV charging solution in Australia.

Genomic assessment of early signatures of adaptation from a provenance trial

This project performs genomic analyses of juvenile eucalyptus trees from the Collaborator’s climate-adapted provenance trial at Nardoo Hills. The expected outcome is to identify genetic variation associated with greater fitness in two key revegetation eucalypts, Eucalyptus microcarpa and Eucalyptus melliodora. The project may provide valuable insights into the genetic mechanisms underpinning resilience and adaptation, which are vital for informing conservation strategies and restoring ecosystems.

Policy and regulatory aspects of water banking in the Murray-Darling Basin, Australia and California, USA

This project investigates the policies and rules that influence the ability and motivations to store river water in underground aquifers for use during drought. The expected outcome is information that can be used to design policies and regulations that allow greater uptake of water banking techniques. These practices can provide cost effective water supplies during drought conditions for Murray-Darling Basin communities, irrigators, First Nations peoples, and the environment.

Enhanced resilience of irrigated agricultural floodplain landscapes of the Murray-Darling Basin

This project investigates the functional response of floodplain vegetation to environmental drivers at multiple scales in the Murray-Darling Basin. The expected outcome includes a series of tools for prioritising the management of floodplain vegetation communities across the landscape of the Murray-Darling Basin at multiple scales and identifying thresholds for environmental watering. The potential benefits are spatial data and new knowledge that will guide future environmental flow management for the benefit of iconic floodplain vegetation communities and related ecosystem services, especially those important to irrigated agricultural, such as water quality. 

Enhancing sustainable land management practices through natural capital profiles

This project addresses knowledge gaps in environmental sustainability and social marketing, focusing on river health and sustainable land management practices via co-creation and adoption of Natural Capital Profiles. The expected outcome is to extend existing frameworks in social marketing, with a particular emphasis on the Co-Create – Build – Engage (CBE) process, to design more user-friendly behaviour change solutions for improved river health and land management practices. The project will potentially increase the adoption of Natural Capital Profiles by landholders, enabling healthier river ecosystems and better land management.

AI-driven automatic translation of blueprints into construction instructions 

This project develops AI models to address the critical gap in automated interpretation of blueprints in construction domain. The expected outcome is an AI system that translates complex blueprints into plain language actional construction instructions. The project outcome will reduce errors, delays and cost in construction industry, enhancing productivity, safety, and sustainability.