Nirajan Shiwakoti: Deep thoughts about surface transport

Nirajan, could you tell us a little about where you work and what you do?

I’m a senior lecturer at the School of Engineering at RMIT University, and I’m the program director of sustainable system engineering program here. I’m also the leader of the urban transportation systems research team. That’s a part of the cyber physical system research group that looks into surface transportation, aviation, and other modes of transport as well. But my main focus is surface transport.

And as a teacher, I teach subjects in intelligent transport system and sustainable transport systems at RMIT.

Just to clarify, in your teaching, what sort of modes of transport, what aspects of transport?

It’s looking at multimodal transportation, from pedestrians, to cyclists, public transport, freight, cars. So teaching the students how we can better accommodate all these transport modes and connect them together. That is, I think, the ultimate goal of any transportation engineer.

When I started my career as a civil engineer, I was looking into transport projects, water engineering, structural design, and so on. And then with transportation, I have always had a passion, maybe because I grew up in a developing country and I could see the road fatalities and all the congestion, and I was always thinking, maybe we can do it in a better way.

That was in Nepal?

Yes. And then I went to Japan in 2004, and stayed for two years doing my Masters degree. And that’s where I got introduced to the field of intelligent transport systems (ITS). Japan I think was, and indeed still is, probably one of the leading countries in terms of ITS technologies. And that’s where my interest and passion began.

Japan also allowed me to see how the effects of unplanned disruption on the transport system, with the country prone to earthquakes that, resulting not only in gridlock, but also evacuation procedures.

And that’s where I started initially my work on that area, looking more on the human factor side of it, evacuations and managing the large number of passenger crowds in major transport hubs like train stations and airports.

Just imagine if something major emergencies or disasters happened in the Melbourne CBD. I’m not sure if we have a proper plan to know where exactly our refuge points are, or should be.

Does that surprise you given recent history?

Yes. I think we need to be prepared for these sorts of emergencies. Before the September 11 attacks I don’t think there was that much concern about the security of the transportation system. Now you can see in Europe and in the US the systems are always on high alert. I think it’s time that we also start thinking about it and make our system more resilient.

So that’s how I started my transportation career and I’ve been working for a long time in the areas of vulnerable road users, and public transport use. And what I realised is that people are the critical element of the transportation system, because if you don’t have proper understanding of how humans behave then you’re never going to solve transport problems.

And there are many instances where we engineers thought our ideas were going to work, but we forget about the people, and that’s where major projects can fail to deliver on their promise.

So you think data is key in delivering on the promise?

Yes, but we also need to properly understand the human behaviour side of it as well. I’ll give you an example. In the early 1950s and 60s when traffic flow models were developed, it was all based on hydrodynamic equations. What that means is they were treating traffic flow like the movement of fluids. And still they are using this equation.

But what they forget is that there’s a person behind the wheel, and that makes all the difference. They don’t behave like a fluid, they’re making decisions, and each person has their own set of cognitive abilities. So, unless you have that understanding, these hydrodynamic equations that consider traffic flows as a fluid, I think that’s a huge problem.

It’s dangerously theoretical.

It is. And to be honest, traffic engineering is a young field. It’s not like structural engineering which has been around for a long, long time. Being a new field that’s why we have been borrowing theories from the other fields. But I think now we are realising the human factor in engineering is also important in terms of understanding how traffic flows.

Yes, on reflection it’s odd it took us this long to get around to that! Where did you do your undergrad and post grad degrees?

I did my undergrad in civil engineering in Nepal. After working for a few years, I went to Japan did my Masters, and then I came to Australia and did my PhD at Monash University.

What year was that in Australia?

It was in 2007. And since then, I’ve been working in this field, but I have always had a passion in the area of transport. Even when I worked as a civil engineer, I was involved in geometric designs of roads, and infrastructure development.

It seems a somewhat common theme with people I’ve interviewed for this series, that they’ve studied in Japan and come away with the ITS bug. So, of your work since you specialised in ITS, what is it you’re most proud of?

The one project that I’m really proud of is when I looked at how we can use swarm intelligence to understand the management of the crowds, or large numbers of passengers. I conducted what I think is a pioneering piece of work in terms of using animal models to understand human traffic congestion.

Urban traffic congestion is somewhat new to us, it’s been with us for perhaps past 50 or 60 years. But organisms like ants have been dealing with their congestion for over millions of years. So it stands to reason that they could have teach us something. If you ever happen to see in your gardens the way ants move, if you look from the top it can look like cars moving on highway. But you’ll notice that ant movements don’t result in congestion. They move smoothly.

Then why are they behaving like this? Because they don’t have the selfish element in it, they work for the benefit of all. But if you look at the way we move in traffic, there’s a selfish element in us. We want to go to our destination as quickly as possible, and that desire to get to our destination as quickly as possible makes the traffic slower, something called as ‘faster-is-slower’ effect. If we could slow down a little, and maintain a certain constant gap within the traffic, that probably would help the traffic to move much quicker, and more smoothly.

Recently a researcher demonstrated this. They did a real test on one of the busiest freeways in Tokyo, and found that there only needs to be eight vehicles who move at a slower pace (than other cars) that can have a significant effect on traffic movement on the freeway. A distance between cars of around a 40-metre gap was maintained in the traffic, to eliminate bunching due to constant breaking. But if you try to maintain a constant gap, while you might move more slowly, you will reach your destination more quickly.

It’s a tough job, not only do you have to engineer transport, you need to engineer human behaviour! Your work on insect behaviour and congestion, has anything been done with it?

Oh yes, it has been submitted and published, and already several PhD students in Australia are following my methodology. They are using ants as animal models especially for the management of the passengers’ movement, and particularly evacuation congestion.

We also worked with Public Transport Victoria in an ARC Linkage project where we utilised the knowledge derived from my research to manage passenger crowds in major stations in Melbourne, like Richmond, Southern Cross and Flinders stations.

These tests run in an experimental mode, because as you can imagine evacuation is a real event, especially a panic situation. You can’t light a fire in a stadium and see how people behave, instead you need some set of empirical data model. Ants provide that opportunity for observation, so it has more life than say an equation.

Are animals other than ants used in these types of experiments?

Since my work was published, in 2009, some researchers in Europe and China have used bigger animals, such as mice and sheep. And the findings are quite consistent in terms of the general underlying dynamics and collective behaviour.

That said, we do need to be careful as well, it’s not a simple substitution of animal behaviour for human. It’s more about understanding the common underlying dynamics of how collective behaviour are influenced by architectural configurations of escape area and examining the clogging transition at bottlenecks.

Okay, moving along It’s hypothetical time. Someone’s come to you, wants you to conduct a project, be it study or actual project, with an unlimited budget and a generous period in which to conduct it. What would you like to do to make a big impact?

I think as a transport engineer, I have to be honest with myself, and I think unless we control our urban sprawl, I think that’s the major issue at the moment. So we need to control the urban sprawl first. If we keep allowing the sprawl, then whatever efforts we put in is a band aid approach. You might fix small problems, but there’s bigger problem happening.

And how would you do this?

That’s a tricky question. As transport affects different aspects of our lives, we need to bring a wide variety of experts together – urban planners, urban designers, transport engineers, computer scientists, and more. All of these experts should come together, bur really such gatherings don’t really happen. Yet.

We see computer scientists predicting traffic flow without knowing the theories behind how the traffic works. And then the transport engineers. They don’t acknowledge the computational capabilities that we have now, that can be used to solve problems. And then there are the urban planners who come and proffer some ideas, but again, they may not understand the other dimensions of the traffic problems. Unless we bring all these people together, then we can’t have a winning solution.

I can tell you a couple of very interesting, revealing statistics. I think this study was done way back in 1999 in US, if I remember correctly, where researchers plotted a graph showing the percentage of Gross Regional Product spent on passenger transport versus road availability and rail orientation. The study found that cities with the most roads (e.g. US cities, Australian cities) are the most costly, and the most rail-oriented cities (e.g. wealthy Asian cities like Singapore, Tokyo, Hong Kong) are the least costly.

Further, it was observed that lower the density of the city, the more it costs to operate passenger transport systems. Providing roads and facilitating cars are the elementary mechanisms for sprawling a city and is an expensive way to build and maintain a city. Therefore, by controlling sprawling and focusing more on re-urbanising (e.g. transit-oriented developments) would ensure savings in money which can then be invested in other productive and innovative aspects of the city.

You mentioned before quite a few professions to come in and work together. You left the arts. Do you think areas such as sociology and artists have a part to play?

We need all these people, and more! Let me give you another example, some fantastic work done by the people from sociologists and psychologists. Have you heard about the Invisible Gorilla experiment?

In passing …

Six people were brought in for video observation by a viewing group. Three people of the six were wearing a white dress, and the other three people were wearing a black dress, and they asked them to pass the ball. And what they asked the viewers of the experiment to do was count how many times the people with the white clothes pass the ball to the people with the black clothes.

During this, someone with a black gorilla costume comes there and says hi to the camera and just passes through. But what they observed was that 50% of the observers didn’t see that gorilla at all, they were so focused on counting the number of passes. So that was the Invisible Gorilla experiment.

Now the relevance of that to our traffic system. The experiment showed that people who notice the gorilla are better able to focus attention. However almost half of the participants did not notice it. You can imagine that if you’re driving in Australian major arterials roads, drivers are more used to paying attention to the cars and other vehicles (as we have car-oriented cities) rather than cyclists/motorcyclists or to some extent pedestrians. Cyclists and pedestrians in a way represent that ‘invisible gorilla’ and only drivers with better control over attention would be more likely to notice those ‘unexpected’ road users.

Therefore, some people are susceptible to inattention blindness and others are not, which would then have implications in the way we design our road traffic systems. We need to acknowledge the limitations of human memory and design accordingly a safer road traffic system for all road users.

So, when there is a complex problem, you need to think outside of the box, go beyond your own discipline and see what other disciplines can teach us.

To be open-minded. All right. Part two of the hypothetical, this time you’ve got quite a limited budget and time frame. You need to spend wisely in order to make a big impact on maybe a small thing but make a big change. What would you do?

Probably local councils starting to develop principle pedestrian networks, and principal bicycle networks, and make those connections to public transport and the shopping centres. We see a lot of connectivity issues at the moment, and in terms of infrastructure like that there’s not much incentive from the urban planning and design sides to drive behavioural change in transport. It’s so monotonous to walk as well, there’s not enough gardens and parklands that walkways pass by and through. This sort of improvement to connections can be done at a local level.

I only live two kilometres away from a train station, but if I walk or cycle there I feel a disconnect. I have to walk through a grassland, there’s no proper footpath or bicycle path. So I don’t walk or cycle, instead I take my car. But if there were better walking/cycling facilities I would keep my car at home. These smaller things can make a big difference I think.

I agree and I think you’ve said a few things that a few other people I’ve interviewed have also said. So, you’ve worked in quite a few areas, in the entire gamut of smart mobility, is there something that you haven’t worked in yet that you’d like to?

Recently I started working on the virtual reality and augmented reality, trying to see how we can make use of these technologies to solve some complex problems. We’re working with the Department of Defence, developing a virtual reality system where we not only have a virtual world, but one in which avatars are generated inside the system. So you’re immersed in that virtual world, and then depending on your action, the avatars would also have their reactions. That’s some very complicated programming!

If we can develop this, and right now we’re doing small prototype studies, it would have a lot of applications not only in transport, but in defence, evacuations, etc. If we can get this right, then I think it would have big implications.

So we are doing that project at the moment plus another for the Rail Manufacturing CRC, looking at how we manage this large number of crowds in major transport hubs. The ultimate goal is to develop a real-time passenger information system that not only manages day-to-day activities, but would also advise what to do in cases of emergency evacuations. What ways we help evacuations take place in a quicker time and avoid bottlenecks. All these strategies can be provided in real-time.

We’re also making use of technology like wi-fi and Bluetooth detections that are not, at this stage, accurate enough but hopefully will be so in the future.

What in the technology area of smart mobility in the next three to five years are you most excited about?

Honestly, I’m excited most about the race toward AV vehicles connected vehicles. I’m not worried about how they might solve traffic problems but I’m mostly interested in the outcomes from the race. I have this feeling that maybe there might be something, a breakthrough that can have applications not only in transport, but also in other fields.

You know the story of Frankenstein, yes?

I know the book, and was terrified by the films as a kid.

There were some crazy experiments done in the late 17th Century and early 18th Century, where researchers gave corpses electric shocks. They observed that corpses’ hands and feet moved as a result of the shock. Researchers thought at the time we could reanimate corpses. Of course they came to realise that this result was not possible.

But later you can see outcomes from those crazy experiments. For instance, defibrillators were invented. A no to bringing the dead back to life, but instead a pathway to equipment that could save lives.

I see a similar situation to this. Everyone is pitching AVs and CAVs like it’s going to solve our traffic problems … I think this is a mistake. It may solve certain aspects, it might save some lives and prevent road accidents, but it’s not going to solve all the traffic problems that we have. I’m more excited more benefits we don’t know about yet are on the way.