Andrew Higgins: In TraNSIT

The man steering TraNSIT, a nifty online tool to help freight operators work out the most cost-effective and carbon-emissions reducing routes, is Andrew Higgins is a Senior Principal Research Scientist at CSIRO, Australia’s main science agency.

The Transport Network Strategic Investment Tool is now seven years strong, continually improving and expanding plus helping governments make better decisions. The freight industry needs it because of the long supply chains and sometimes thousand-kilometre plus distances between production, processing and markets in Australia.

What’s your day-to-day work like?

It’s very applied. I’m always out on the ground, meeting industry. Out of all of the tools for industry I’ve done in the past 25 years, TraNSIT has by far had the biggest impact and demand by government and industry.

I completed my PhD in 1996 from the Queensland University of Technology, then joined the CSIRO. My research interests include modelling and optimising complex industry and infrastructure systems. Rail transport planning and logistics, intermodal transport and agricultural supply chains are some of my focus areas. I’ve been involved in about 70 projects, have published more than 60 articles in scientific journals plus delivered 55 conference papers.

I’m mid-career now and my focus is a bit more removed from the scientific coding and equations to focusing on high to turn science into high impacts for industry and government.

What exactly is TraNSIT and how does it work?

TraNSIT is a spatial computer model for simulating transport cost benefits from infrastructure investments (road upgrades, use of rail versus road, processing and storage facilities) and policy interventions in agriculture logistics. It got off the ground thanks to funding from major Australian Government initiatives.

The model maps millions of vehicle trips across thousands of supply chains between production and domestic and export markets. On a scale never seen before, it’s brought together large data sets from more than 400,000 enterprises and about 550,000 supply chain paths across the agriculture sector – that’s 98% of the sector. It covers 60% of non-urban road freight and 95% of rail freight.

Government and industry commission CSIRO to use the model to put in their needs to identify transport related investments with the greatest bang for buck. We estimate it can reduce transport cost by 70% more than without TraNSIT. Putting that into perspective, it currently costs $10 billion to transport Australia’s agriculture and forestry via road and rail each year.

Each time we run the model, it produces up to 10 terabytes of output and that’s just for Australia.

Actually, we’re also using other information such as artificial intelligence to get a better understanding of behaviour at truck stops and likely trips that tourists would take.

How easy was it to source and use all that data from commercial operators for TraNSIT?

It’s taken long-term relationship building with industry to get the data. To give you an idea of the scope, it comes from more than 200 companies, industry associations, or government agencies.

We’re using data we don’t own, so we haven’t gone out for a shopping list of data that goes beyond what we need. We’ve asked for specific information and been really clear with industry organisations about how it’s going to be used. So, it’s confidential data, but we’ve desensitised it, so you can’t see the information at an individual enterprise level.

We’ve built the model from the ground up and are using our own computer code. The back end was developed in Australia and remains housed in Australia. We also had to give key stakeholders reassurance that the data won’t be distributed and it will be secure.

Most of our presentations are to industry and government stakeholders through round tables or major industry symposiums. Through these activities, we discuss TraNSIT’s benefits and work through how it can be used on the ground to inform infrastructure investment. We have to take a plain language approach to show benefits in terms of dollar saving per tonne when they use TraNSIT to select a freight route.

TraNSIT emerged to help the agriculture industry – was that your background?

No, my background was in operational research and rail transport planning in the early 1990s. Since the mid 1990s, my technical and research work has always been linked to industry and most of my career has been in the agricultural and logistics space.

Is the model being used outside of Australia?

Yes, it’s a tool that can be adapted to domestic or to international markets. We’ve even customised it to major industries in in Indonesia, Vietnam and Laos.

A challenge is that we don’t own the data, so commercialising the Australian product is difficult. Using TraNSIT in a larger international sense would require industry or key players in those countries to work with use to build a new one for that sector. We have a more generic international version that has the functionality for vehicle types, trips, etc.

Using the road network in those countries allows us to transfer it and be populated. To do this, we have to build individual project teams in those countries, so we’re not trying to do it all from Australia. When you’re mapping a whole supply chain for a country, there is a need for the in-country partners to take ownership of the tool and eventually use their own data independently of the team in Australia.

What are you working on now?

Over the past six months, we’ve expanded TraNSIT to cover tourism to give us a better understanding of the roads tourists travel, be they domestic or overseas visitors. I’m leading this project which aims to give an overall picture of how road upgrades – infrastructure investments – will benefit the tourism sector. A key factor is if you make an improvement in the road network, that makes it more attractive to tourists to visit particular destinations en route.

The Federal Department of Industry, Innovation and Science is co-funding this project through the Office of Northern Australia and the Department of Infrastructure, Regional Development and Cities plus AUSTRADE. We’re focusing on case studies of Northern Australia. We’re just wrapping up the study and expect to release the results soon.

To finish, here’s a couple of hypotheticals for you. If money and time weren’t an issue, what project would you tackle to improve transport? No time limit, and unlimited funds?

I’d look at exploring how to come up with a complete road, rail, air and sea network that’s truly integrated and built for long term adaptability to freight and domestic travel in Australia. Through high performance computing, it would interactively use the supply chain information from all vehicle drivers’ trips, along with complete data sets of vehicle fleets and transport networks. That would all be updated in real time.

TraNSIT would be used to inform that. It would take in long-term data – freight movements for the whole year and origin-to-destination information so operators can plan better for the longer-term supply chain.

It would also help identify where investments would be needed across the networks to get the best integration, the fastest freight flow and lowest costs to operators. The integrated network would be resilient to climate impacts such as floods and fire that disrupt part of the network. It would have the capacity to suggest alternative routes/modes in real time and identify tactical logistics and investments to recover as soon as possible.

My focus is freight, but it would have spinoff benefits for domestic road transport use, commuting and tourism traffic.

OK, and if you had a limited budget and time frame to do a project, but still have a decent impact – what would work for you in this scenario?

Let’s say, we have five years to do this. We’d love to have a version of TraNSIT that is automatically updated with transport networks and freight data for all commodities. This includes a means where every company can safely provide desensitised freight information for TraNSIT. They’d provide their freight volume loads, origin-destination, and mode, as well as soft data such as regulated considerations and bottlenecks …

Such a tool would provide a fast response to informing infrastructure investments, disaster recovery plans and seasonal logistics planning using the most up to date data available. It would be an adaptable TraNSIT model that takes into account progressively changing freight.