Professor Vinayak Dixit uses AI tools across all aspects of his work, including teaching, research and from a global perspective.
Professor Vinayak Dixit has a dual role at UNSW. He is the Associate Vice-President Global Research and Innovation in the University Office of Global Affairs and a Professor of Transport Systems in the School of Civil and Environmental Engineering.
From a teaching perspective, the first time Prof. Dixit used AI was in his course on human factors in engineering. It was a relatively new course with no previous exam papers, which were the first thing his students wanted access to. Prof. Dixit suggested they use AI to help.
“I gave my fourth-year students the content and recommended they put it into their choice of AI tool and have it ask them questions about the content for them to try and answer. The amount of trust the students placed in AI tools was interesting and impressive at the same time,” said Prof. Dixit.
There were two approaches to how his students could use AI. One way was telling them to use AI themselves, the other way was to use it with them in a tutorial setting. He preferred the latter approach.
“Students would present their answers to me, but we found the answers were only getting about 20-30% of the content right at that time. What it was missing was the rationale and argument and keeping the answers clear and concise,” said Prof. Dixit.
“We focused on the interpretation, finding the content you could trust from AI and then asked it questions to evaluate the responses.
“It was a really interesting exercise we did together. It helped me personally as well and informed how I could use AI in my own research.”
AI sandbox for research
In terms of research, Prof. Dixit started using AI to cross-check his work.
“If AI doesn’t understand, it says it’s wrong,” said Prof. Dixit. “You have to get it to realise what you’re trying to do and then set up your own sandbox to start testing it.”
Prof. Dixit has also found AI really good at developing proofs, particularly around mathematics and other areas of literature.
“The output may not be intelligible or understandable, but it gives you directions of inspiration. For me that was interesting, but you have to prompt it in an appropriate way otherwise it can take you down a rabbit hole,” said Prof. Dixit.
In research there is so much data available already, but for Prof. Dixit it’s not only about using AI on the data but also for finding data.
“Whenever somebody publishes a paper, they include the data,” said Prof. Dixit. “So AI actually helps inform what data is available and you can use it to test your theory pretty quickly.
“It’s also helpful for identifying new data sets for experiments where you might not yet have the data.”
Transport system digital twins
AI is being used in the transport systems space but not to the same extent as other fields.
“Transport systems is a very traditional field,” said Prof. Dixit. “In my experience, when dealing with more physical material things, it takes longer for digital technologies to catch up.”
One way Prof. Dixit has been using AI is creating digital twins through virtual reality. A digital twin is a digital version of the physical world we interact with. It’s used to evaluate behaviours which in turn informs real-world design.
“We use it to assess behaviour and designs around things like roads, cycleways and pedestrian crossings. The digital version enables us to manipulate and test what impacts it could have,” said Prof. Dixit.
“For example, if we are planning a new road, it can help us assess the impacts on congestion or how it feels from a design perspective.
“It’s also very useful when we have very complex data. Data streams need to be tagged, collated, structured and analysed. The digital twins save us a lot of time throughout this entire process.”
Enabling startups to grow globally
In Prof. Dixit’s current role in Global Research and Innovation, his focus is to elevate opportunities in research and innovation for UNSW researchers, startups and spinouts in a global environment. He has found AI particularly useful for SWOT (strengths, weaknesses, opportunities and threats) analysis, identifying opportunities and risks.
“I pull in evidence to inform the hypothesis and business case,” said Prof. Dixit. “All the heavy lifting of searching, researching and rapidly testing those business cases and doing opportunity and threat assessments has been very useful.”
One of the programs he helped set up in collaboration with the Division of Research & Enterprise is the Global Landing Pads Awards, now called the Global Innovation Foundry. Working closely with UNSW Founders, the program supports startups to grow globally.
“AI has enabled us to assess opportunities and risks for startups and technologies in different regions,” said Prof. Dixit.
“It helps us narrow down the questions we need to ask and identify places we can support our academics and startups to grow in different regions.
“Even with the feedback we can give startups and our innovation and research teams, it’s helpful. Also whether it aligns in terms of our vision, goals and philosophy. It really helps us build our business case and frees up our time to focus on implementation.”
As a final note, when using AI in such varied ways, Prof. Dixit wants to highlight the importance of responsibly and safely using the tools, particularly when inputting data. It needs to be cautiously managed.
“We need to think about how we’re using AI and the things that could open us up to risk later,” said Prof. Dixit. “If we’re not sure, it’s important to ask questions. With AI, the stakes are different. Once it’s there, it’s permanent and the damage is done.”
- Log in to post comments