Meet UNSW Aviation’s Associate Professor Richard Wu. With a background in engineering and aviation, Richard’s research focuses on aviation operations using mathematical models such as optimisation, machine learning, simulation models, and now AI to solve issues and challenges.
My Bachelor's degree is in Civil Engineering. When I did my Master's degree, I focused on Transport Engineering. I further focused on air transport study in my PhD and have been working on aviation since my PhD in 2001. I grew up in Taiwan and worked for Taipei City Government and Taipei Songsang International Airport before starting my academic career with UNSW Aviation. I enjoy teaching and research. Outside work, I'm a keen gardener, a ‘fix-everything’ Dad with two children, and a triathlete. I finished Ironman races three times in the past five years, including New Zealand Ironman and Port Macquarie Ironman.
My research focuses on the operational side of the aviation industry. This includes airline operations, airline resources optimisation, airport operations, airport terminal space planning and airport retail management. In addition, I also study passenger behaviour such as ticket purchase behaviour (airline revenue management & pricing), passenger retail purchases at the airport and passenger purchase behaviour in relation to tourism. Most of my work is technical and involves mathematical models such as optimisation, machine learning, and simulation models. I have recently used artificial intelligence (AI) models to solve aircraft scheduling problems.
In my career, I've focused on the operational side of the industry and the customer side of the aviation business. This line of work brings me to see the operational side of the aviation industry. I guess that’s why I have many aviation photos in my photo collections. I have had opportunities to work with IATA, airport operators, airlines and consultants. The commercial side of my work leads me to work with the retail industry, tourism partners and different types of technologies such as real-time positioning of a mobile device via WiFi and Bluetooth networks and cloud computing.
I've recently worked with airlines and airports, including Qantas, Virgin Australia, Royal Flying Doctor Service, and Taipei Songshang International Airport. I also worked with London City Airport, Gulf Air, British Airways, EasyJet and a number of investment banks and fund managers.
I'm currently working with the Royal Flying Doctor Services (NSW) to enhance their pilot rostering and aircraft utilisation. I am involved in a project with uDASH (the Big Data Analytics Centre of UNSW) and UNOVA (a UNSW Research Centre for digital transformation in the Business School) working on transportation job bidding data from Kent Removals. This data analytics project extends my work on airline resource management to logistic transport fleet management. I'm also working with a start-up company, H2ec, a hydrogen energy company. Our recent collaboration focuses on transforming regional aircraft to hydrogen-powered green engines and studying how aviation infrastructure should be enhanced to accommodate this future demand.
The media often approach me for commenting on current affairs in aviation. These topics mostly focus on airline strategies, flight operations (delays), flight disruptions and incidents. I was on SBS in 2022 talking about high ticket prices after travel restrictions eased early last year. I was on ABC Radio in late 2022 talking about the airport labour shortage issue and why baggage handling was challenging then. Last week, I was interviewed by the AGE newspaper about the change of passenger boarding procedures by Qantas and how this may affect Qantas operations and passenger experience.
Industry practitioners also approach me on topics like flight scheduling issues, crew scheduling, flight delays, and data modelling. These topics are relevant to my research, and these are great opportunities to build collaborative relationships with industry partners.
The current flight delay coding system in the Ground Handling Manual of IATA was based on my research published in 2014. In the past, flight delay data was collected but the quality was poor and did not help data modelling. This new delay data coding system greatly enhances the data quality that airlines collect. The enhanced quality allows deep data modelling and analysis, providing valuable insights into airline scheduling and operations.
I invented an aircraft turnaround monitor system by using mobile devices in 2006. This system was later commercialised by a British company (Avtura) in 2007 and SITA in 2009 to assist ground operation data collection and real-time monitoring. UNSW didn't keep the patent, so the opportunity was lost. Avtura has been developing this field in aviation since 2006, and I'm collaborating with this company on big data analytics.
I am currently working on an AI-enhanced aircraft maintenance scheduling project with my PhD student. Since aircraft maintenance is essential for flight safety, maintenance scheduling is key to ensuring flight safety. Maintenance jobs can be done at multiple airports, and some jobs can be done without a hanger when an aircraft is on the ground. Optimisation models can do scheduling, but building such a model is difficult and solving the optimisation problem is often mathematically difficult. My team is trying to use Reinforcement Learning (AI models) to solve this scheduling problem. Preliminary results are promising; we can solve the same problem with a solution as good as optimisation models.
I bring my research to my classroom. Given the fact that there was limited teaching material in airline operations, I wrote my own textbook in 2010 by bringing in my research outputs, literature and industry practices. When contents are too technical for undergraduate students to understand, I will teach students manual ways of doing the same job by learning research outcomes from the literature and combining them with industry practices. Classroom learning with a small industry problem is beneficial for students because the skills student acquire are relevant to real industry problems.
I will focus on big data analytics and AI-enhanced scheduling in the near future. Airlines are collecting a huge amount of data, but data alone doesn't help decision-making. Data modelling will play a key role in shaping future airline business strategies and operations. AI models will be helpful in airline resource allocation and operation management. Given the flexibility of AI models and quick response time, AI models will become essential helping tools for scheduling, revenue management (pricing), and operation management (flight disruption recovery). In addition to airlines, airports will also benefit from the development of these two major areas.