Milad Mousavi Wins Best Presentation Award at ICMLT 2025

 

Milad Mousavi, a RIIS Hub student and researcher from the School of Civil and Environmental Engineering at UNSW Sydney, was honored with the Best Presentation Award at the 10th International Conference on Machine Learning Technologies (ICMLT 2025), held in Helsinki, Finland from May 23–25, 2025. His paper, “Online Deep Transfer Learning and Multi-Sensor Analysis for Enhanced Underground Monitoring,” introduces a novel approach to improving safety and efficiency in underground environments.

Tackling the Challenges of Underground Monitoring

Underground operations, such as coal mining, face significant challenges due to complex spatial layouts, unpredictable geological conditions, and hazardous gas emissions—particularly methane. Traditional monitoring systems often struggle with data scarcity and adaptability, especially during the early stages of operation.
Milad’s research addresses these limitations by integrating multi-sensor analysis with online transfer learning, enabling real-time prediction and monitoring of methane concentrations. His model leverages a Long Short-Term Memory (LSTM) neural network trained on data from a Polish coal mine and fine-tuned using online learning with data from a Chinese coal mine.

Key Innovations and Results

The study demonstrates that combining historical data with real-time updates significantly enhances prediction accuracy. Notably:

– The online transfer learning model achieved an R² score of 0.93, indicating high reliability in methane concentration predictions.
– The model trained with transfer learning showed faster performance improvements compared to models trained from scratch.
– This approach helps mitigate the “cold start” problem in new underground projects, where data is initially limited.
By continuously adapting to new data, the model supports timely interventions, reduces false alarms, and improves operational efficiency—ultimately contributing to safer underground environments.

A Vision for the Future

Milad’s work, as a RIIS Hub student supported by the ARC Research Hub for Resilient and Intelligent Infrastructure Systems (RIIS), sets a new benchmark for intelligent underground monitoring. His presentation at ICMLT 2025 not only showcased technical excellence but also highlighted the practical impact of machine learning in infrastructure resilience and safety.
As underground operations become increasingly digitized, innovations like Milad’s will play a crucial role in shaping the future of smart, adaptive monitoring systems.

RIIS/ CIES 3MT competition

We’re very proud of students who have been able explain their complex research simply, in less than 3 minutes, with no notes and using only one slide ….and wow a crowd while competing in the UNSW Civil and Environmental Engineering #3MT competition (Heat), hosted by The Centre for Infrastructure Engineering and Safety (#CIES).

Among them were:

🔵 Higher Degree Researcher/PhD student Kartika Nur Rahma Putri spoke to her work in Automated Material Management in Construction

 

🔵 Higher Degree Researcher (#HDR)/PhD student Yiyang Liu spoke to his work developing Rubberized Geopolymer Concrete Pipe Using Dry-Cast Technology

 

We congratulate event winners, and wish them luck as they progress into the next round. We also acknowledge the efforts of all students taking part in 3MT competitions. It’s inspiring to see how passionate students are for creating change during their three-minute #PhD Thesis introductions, where they explain their research in simple terms to an audience and panel of judges, including how their research is unique, why it is important and how it could be used to improve the world 👏

Some of the other 3MT competitors and their UNSW and RIIS Hub: Resilient and intelligent infrastructure systems colleagues, including Heat competition judge, RIIS Hub Chief Investigator Mehri Makki Alamdari and event Emcee RIIS Hub Chief Investigator Johnson Xuesong Shen.

 

 

 

 

Winners of the 3MT Competition:

1st Place: Tareq Khondoker

2nd Place: MD. Ibrahim Mostazid

3rd Place: Gunja Shah

4th Place: Mikael Gilbert

They will advance to represent the centre in the School Round on Friday, 6 June 2025. We wish them the very best!

The Centre for Infrastructure Engineering and Safety (CIES) is a research centre focused on understanding and developing ways to investigate and predict the behaviour of #engineering #infrastructure under a range of conditions. Located at the UNSW Australia’s School of Civil and Environmental Engineering, CIES projects incorporate a number of engineering disciplines – from structural engineering to geotechnical engineering, and engineering materials to computational mechanics

RIIS VIsitor – Prof Alexander Scheuermann from the School of Civil Engineering at UQ

We were privileged to have a seminar presented to RIIS Hub members at UNSW Sydney on Friday 14th February 2025.

Prof Alexander Scheuermann is a recognised expert in geotechnical and environmental engineering renowned for his expertise in the development, modification, and application of electrical and electromagnetic measurement methods for various applications. Since 2010, he has held a position at the School of Civil Engineering at the University of Queensland, where he heads the GeoSystems Group. This group is dedicated to investigating complex multiphysical processes in porous media across the scales, including processes under unsaturated conditions. His academic journey commenced at the Karlsruhe Institute of Technology, Germany, where he completed his Habilitation Thesis on “Spatial Time Domain Reflectometry (Spatial TDR) in Geohydraulics and Geomechanics”. For the advancement of the Spatial TDR technology for key industries in Queensland he was awarded the Queensland Smart Futures Fellowship in 2012, and he was invited to join the Technical Committee of the International Society for Electromagnetic Aquametry (ISEMA) for his contributions to this research field. He was awarded the prestigious ARC Future Fellowship in 2018 on the topic of erosion. Prof Scheuermann is Editor of Acta Geotechnica and Associate Editor of the Journal of Geophysics and Engineering.

Publications

RIIS EOY Poster Competition

RIIS Hub students took part in a Poster competition during our end of year workshop, with judges ranking entrants on 2 key qualities:
* Effective communication when describing their research: Actively engagement with the audience, inviting questions and spurring discussions
* Poster is visually attractive, well-balanced with text and graphics enhancing the understanding of the content

Winning Poster: Vladislav Alekseev
Research Poster Title: Seismic Data Denoising with a Single-Shot Deep Learning Approach: Double-DIP
Research Insight:
Denoising is an essential step in seismic data interpretation. We proposed a novel deep learning approach where one neural network predicts the clean signal while another estimates the original noise. This single-shot method requires no additional training data, working directly on a single image. It effectively preserves structural information while achieving superior denoising performance.
Industry partner on this project: Kumul Petroleum

Runner Up Poster: Amir Fila
Research Poster Title: A Graph Theory Approach for Aerial Point Cloud Segmentation and Processing
Research insight:
The majority of research on aerial point cloud data segmentation uses the term ‘complexity’ vaguely by abstractly assigning higher complexity to those point cloud datasets with higher computation requirements. As part of my PhD research, I aim to provide a mathematical approach, leveraging graph theory and Laplacian matrix analysis, to address the complexity elements inherent in aerial LiDAR captures. This approach will serve as a valuable tool for understanding complexity across different sections of a given point cloud dataset, ultimately aiding in noise recognition and boundary detection processes.
Industry partner on this project: Woolpert Asia-Pacific / Woolpert

Congratulations both on your work. We also thank our industry partners for supporting RIIS Hub research, practical research application and educational opportunities.

Vlad and Amir are pictured with their posters as they are congratulated by RIIS Hub colleagues: Hub Director Prof. Nasser Khalili, Hub Deputy Director Prof. Abbas Rajabifard, and Hub Chief Investigators Prof. Sisi Zlatanova and Prof. Bijan Samali.

2024 RIIS End of Year Wrap-up

The RIIS Hub held their 2024 wrap up event at The Coogee Bay Hotel, Sydney.

The Director and Deputy Director of our Research Institute started the day with a warm welcome and an update on our latest research achievements. This set the stage for a day filled with insightful presentations and networking opportunities.

We were pleased to introduce key members of our team and industry partners:

  • Garry Mostyn: Chair of the Industry Advisory Board, who gave a summary of his expertise and what he will contribute to the future research direction of the hub.
  • New Chief Investigators and Researchers who gave brief presentations on their research interests and the projects they would be participating in in 2025

                     Arman Khoshghalb – Associate Professor – UNSW Sydney

                    Jing Zhao –  ARC DECRA Fellow – UNSW Sydney

                    Huadong Mo – Senior Lecturer – UNSW Canberra

Our industry partners shared their latest innovations and research:

                   Dr Govinda Pandey – Rockfield Technologies Australia Pty Ltd – Chief Executive Officer

                   James Linke – GeoAI – Chief Technology Officer & Chief Financial Officer

                   Huadong Mo on behalf of Dr James Ciyu Qin – Advanced United Technologies Pty Ltd – Process Control and System Optimisation Specialist

We stopped for lunch which proved to be a great opportunity to network and discuss the morning’s presentations.

The RIIS poster competition was held from 2pm where Higher Degree Research (HDR) students presented their posters, showcasing their cutting-edge research projects. This session was a highlight of the day, offering a glimpse into the future of our field.  Our thanks to Arman Khoshghalb for convening this session.

The day concluded with A dinner which was a relaxing and enjoyable end to a productive day providing further opportunities for networking and collaboration.

RIIS Hub is working to minimise damage to homes when disaster strikes

Australia is a land of ‘droughts and flooding rains’*, and when they occur soils can swell or shrink, which can result in damage to buildings and homes, and pose a risk to human life.

The personal and collective financial cost of addressing flood-related structural damage to homes is not limited to Australia. It is an international challenge.
Damage to infrastructure could be minimised or avoided with a better understanding of soil behaviour under different conditions, and RIIS Hub is contributing to global learning on the topic.
To further explore international learnings on the topic, RIIS Hub team recently welcomed eminent scholar and researcher in soil-related structural impacts Associate Professor Mathilde Morvan from the Pascal Institute & POLYTECH Clermont in France to explore potential solutions.
A/Prof Morvan’s current research seeks to develop a better understanding of soil characterisation. She is also looking to develop new, inexpensive diagnostic methods that can predict swelling soils near detached houses.
A/Prof. Morvan’s work has close alignment to some of the and predictive techniques being developed in RIIS Hub’s Theme 3 Research program, where researchers are exploring the effect of temperature on the movement of soils using a thermo-plasticity model as part of flaw detection and failure avoidance research.
During her Seminar to the RIIS Hub team, A/Prof Morvan shared

• The cost of damage due to soil shrinkage and swelling on detached houses is the highest among natural disasters in France after floods.
• While methods for characterising these soils exist, they are generally not affordable for most households, and as a result, this foundational soil understanding is often neglected at building and planning development stages.
• One objective of A/Prof. Morvan’s work is to provide a methodology for characterising these soils, and to find simple. Another is to find affordable methods to help spur more due diligence soil investigations to inform structural planning and help prevent damage to homes during disasters.
• A current project proposes to develop a method for monitoring individual houses in high-risk areas, with the aim of providing reliable data.

Following the presentation A/Prof. Morvan spent 1:1 time collaborating with members of the RIIS Hub team, exploring future research and data sharing.
RIIS Hub Director Prof. Nasser Khalili shared:
“We want to make flood affected dwellings are as ‘safe as houses’, and ensure Aussies and their homes are better-prepared to weather future storms.
“Building a collaborative and close-working relationship with other researchers working on related infrastructure resilience projects is a practical way to expand and share knowledge and learnings at scale, and inform the development of more robust technical solutions to solve similar problems.”

Theme 3 RIIS Hub Research involves multiple universities in NSW (UNSW, UNSW Sydney, Western Sydney University) as well as QUT in Queensland. We note that this visit was made possible thanks to NSW Chief Scientist ‘RAAP’ funding.
Theme 3 RIIS Hub researchers are designing and applying:
• Modelling, simulations, and prognostics
• Predictive modelling simulation & performance assessment
• Physics-informed artificial intelligence (AI) and machine learning & explanation

Career and peer-building experiences for RIIS Hub higher degree researchers

RIIS Hub was delighted to support multiple PhD researchers’ participation in the ‘2024 Information Resilience PhD School’ during October in Brisbane, hosted by the ARC Training Centre for Information Resilience (CIRES). 

The event, now in its third year, gathers PhD students from around Australia to take part in a series of activities and updates that are designed to help them build peer-to-peer networks and explore data science practices and career pathways. 

RIIS Hub PhD student Masoud Kamali attended, and was awarded the ‘People’s Choice’ award for his 3-Minute Thesis Presentation 

 “The event workshops and sessions were incredibly informative and engaging, offering valuable insights that significantly broadened my expertise. PhD researchers from various ARC Hubs presented their work across diverse fields, including Generative AI, Data Quality, and Computer Vision, making it a comprehensive and engaging event.” 

 “My 3MT presentation focused on Zero-shot 3D Object Detection in complex industrial environments. This research involves designing and developing a monocular 3D object detection approach for industrial assets, leveraging vision-language models (VLMs), computer vision, and deep learning algorithms.” 

 RIIS Hub PhD student Ensiyeh Javaherian Pour was a Group Competition winner: Data Science and Analytics  

 “Attending the PhD school was a wonderful opportunity. I found the experience invaluable.  

It provided me with the chance to engage with other PhD students, learn about their projects, and share insights. Additionally, I thoroughly enjoyed the other sessions, particularly those led by esteemed presenters who discussed advancements in AI technology and data engineering. The event was both comprehensive and engaging, offering a wealth of knowledge and inspiration.” 

 The event featured a competition designed to expand student knowledge:  

“(Our winning group) achievement was particularly gratifying and impressive for RIIS Hub students, showcasing our capabilities and determination.” 

 RIIS Hub PhD student Alireza Haji Heidari attended, and was also a Group Competition winner
“The event was incredibly insightful and informative and valuable experience, allowing me to engage with leading researchers and fellow PhD candidates.”  

“A series of interactive sessions, including keynotes, workshops, panel discussions, 3MT and poster browsing provided deep insights into cutting-edge data science practices and the future of data-driven research.” 

 “The ‘Beyond Research’ day was particularly enriching, as I learned from industry professionals. These interactions have opened new avenues for potential collaborations and opportunities.”  

 “The Information Resilience PhD School 2024 has been a transformative experience. It has not only enhanced my knowledge and skills but also expanded my professional network. I am grateful for the opportunity to be part of such a prestigious event and look forward to applying what I have learned to my ongoing research.” 

 Program highlights included:
 

  • Interactive sessions, writing workshop, industry engagement & entrepreneurship, 3-Minute Thesis (3MT), poster session, awards presentation and more
     

More information on the 2024 PhD Resilience School can be found here: https://cires.org.au/event/information-resilience-phd-school-2024/ 

  

Professor Abbas Rajabifard awarded CBAS Fellowship

Prof Abbas Rajabifard, Deputy Director of RIIS Hub and Director CSDILA, Dept IE, University of Melbourne has been awarded a Fellowship as the highest accolade from the International Research Centre of Big Data for Sustainable Development Goals (CBAS), at the 4th International Forum on Big Data for SDGs, 6-8 Sept 2024 in Beijing. CBAS has formally been established at the 75th United Nations General Assembly on September 22, 2020.

The Fellowship is to recognise global scientific leaders with outstanding achievements and contributions made in the field of promoting sustainable development through technological innovation in supporting the implementation of United Nations SDGs through harnessing science and technology.

There are five key missions CBAS is working on:

  1. Developing SDG big data platform,
  2. Conducting SDG monitoring and evaluation,
  3. Developing SDG Satellites,
  4. Becoming a think-tank for SDGs,
  5. Capacity development for developing countries

 

RIIS Hub extends climate-disaster readiness expertise and research with a new industry partner: Advanced United Technologies (Audtech)

RIIS Hub is very pleased to welcome Audtech as a new industry partner. 

Audtech is an innovative start-up company, operating at the cutting edge of connected data and AI innovation across complex infrastructure system clients. It strives to initiate government and industry change by delivering benefits aligned to the ‘Internet of Things’ (IoT), and where new innovation can improve infrastructure performance, stability and recovery following disaster situations.

Making sense of the ‘big data’ from system sensors in a highly complex system is, …complex!

That’s where RIIS Hub research, and our talented team members come in.

We’re very pleased to introduce new RIIS Hub Chief Investigator, Dr Huadong Mo (UNSW Canberra) who will lead the Audtech research collaboration.
Dr Mo’s has a depth of expertise in exploring ways to enhance the resilience, performance and security of complex systems with learning-based algorithms, primarily in the emerging fields of power and energy systems, cyber-physical systems and manufacturing systems.

Previously a postdoctoral fellow at the Swiss Federal Institute of Technology Zurich, and with academic institutions in China and Hong Kong, Dr Mo joined UNSW Canberra 2019. He is currently the convener of the Systems Engineering Discipline under the School of Systems and Computing. His skills and experience in data collection and analysis, revealing patterns of system evolution against uncertainties, is well aligned to addressing Audtech’s focus on urban planning and infrastructure prognostics and health monitoring of other system.

Learn more about Dr Mo here:

We’re also pleased to introduce our Partner Investigator on the collaboration, UNSW alum and infrastructure/climate disaster specialist Dr James Ciyu Qin, from Audtech.

Dr Qin has a research background relating to process control and system optimisation and modelling for large scale systems, including power systems. He specialises in enhancing the resilience, performance and security of complex systems through robust optimisation techniques, and by informing preventive approaches that minimise the risk of system failure, and increase infrastructure resilience.

Learn more about Dr Qin here:

Both will work on the ‘Building resilience to cyber-physical power systems (CPPS) under climate-related disasters’ RIIS Hub project, involving UNSW Canberra and UNSW Sydney team members.
In welcoming Audtech as a new industry partner, as well as new colleagues, RIIS Hub Director Professor Nasser Khalili said, “applying RIIS Hub scientific expertise to help avoid, or better respond to, infrastructure failure when catastrophe happens. I’m especially pleased to be working closely with our academic and industry peers at Audtech to enhance data management and systems, and as a result, better serve and ensure the safety of people in our community”.

Artificial Intelligence (AI), Deep Learning (DL), and Large Language Models (LLM) Workshop

An insightful workshop was organized by GRID@UNSW and RIIS in collaboration with the Next-Gen Architectural Manufacturing group at UNSW, delving into the potential of Artificial Intelligence (AI), Deep Learning (DL), and Large Language Models (LLM) for semantic enrichment of geospatial data, intelligent queries and complex analytics.

The session kicked off with and insightful plenary by Rob Atkinson from the Open Geospatial Consortium (OGC), describing Digital Twins as a “systems of systems” connected by OGC APIs to enable cross-domain interoperability and enhance complex problem solving.

The session provided PhD students with a platform to engage in discussions and updates on the latest developments. Presenters highlighted how AI tools are poised to significantly enhance their respective research areas, considering the diverse backgrounds of the attendees. Jack Yafei Sun presented his PhD work exploring information extraction for domain knowledge graph construction and Saeed Heidary presented graph-structured information management systems in construction. Jidnyasa Patil described her Scan-to-BIM DL approach and Yu Gao examined Intelligent lifecycle management of construction and demolition waste in complex environments, focusing on integrating multimodal LLM and agents. Binjie Xu explored the use of multimodal LLM to monitor mental matigue of plant and machinery operators and Mahdi Soheyil Fard discussed AI-integrated solutions for decarbonizing the building industry.

For those interested in more details, the presentations can be accessed at grid.unsw.edu.au/presentations.

Presenter Title Video Link
Rob Atkinson (OGC) Building blocks for digital twins https://youtu.be/OqMko40ju4Y
Jack Yafei Sun Information Extraction for Domain Knowledge Graph Construction (a tunnel maintenance use case) https://youtu.be/rAyjs0oW70E
Saeed Heidary Graph-Structured Information Management System in Construction https://youtu.be/Nl8ge_Hbaa
Jidnyasa Patil Scan-to-BIM: A DL approach https://youtu.be/o_92ytyxKV8
Yu Gao: Lifecycle management of construction and demolition waste: Integrating Multimodal LLM and agents https://youtu.be/M_rV5SqFiLA
Binjie Xu: Use of Multimodal LLM to monitor mental fatigue of plant and machinery operators https://youtu.be/CKcUe7naarg
Mahdi Soheyil Fard AI-integrated solutions for decarbonizing the building Industry (a problem statement) https://youtu.be/ihzS5o9yykU
Discussion https://youtu.be/8CtmVRBgibM