InTrans / Jun 29, 2018

Transportation of Tomorrow: The next generation of engineers

Go! Magazine

posted on June 29, 2018

Pranamesh Chakraborty is a graduate student in the Department of Civil, Construction, and Environmental Engineering at Iowa State University (ISU). He is a graduate research assistant at the Institute for Transportation (InTrans) in Ames, Iowa, and applies his gifts as a budding engineer to work on Intelligent Transportation Systems (ITS), applying big data and deep learning techniques.

As a graduate student earning his PhD, Pranamesh is a part of our next generation of engineers equipped with new ideas, technology, and challenges. I sat down with Pranamesh to ask him all about transportation engineering: Where it is, where it’s headed, and how big data and deep learning will shape the future of transportation.

Pranamesh Chakraborty working in the REACTOR Lab at InTrans.
Photo by Hannah Postlethwait

Why do you want to be an engineer? Particularly, what attracted you to transportation engineering?

My story is similar to many other engineers: I found that I was skilled in mathematics, and I realized that I could be very successful as an engineer. But along with mathematics, I liked to do computer coding. That’s what motivated me to choose a field like transportation, which is very inter-disciplinary and has flavors of both mathematics and coding. Along with that, it has wide applications, since each and every one of us are a user of the transportation world. It feels wonderful to do work that can have a real, meaningful impact.

What is your focus in transportation engineering?

I mainly work on ITS and freeway traffic operations. My PhD dissertation is on freeway traffic incident detection from traffic data and CCTV (closed circuit television) cameras. It deals with the combination of applications of big data and deep learning techniques towards faster, reliable detection of traffic incidents. During traffic incidents, every second matters. And our main goal is to detect these incidents as quickly as possible so that the rescue teams can be dispatched quickly to handle the situation, thereby saving lives and reducing risks to other drivers.

How would you define “big data” and “deep learning? How can these techniques be applied to transportation?

Both big data and deep learning are popular buzzwords in science and technology today. Big data broadly refers to the techniques used when handling massive data sources (in order of terabytes or even petabytes), which cannot be handled by traditional single CPUs (computer processing units). It requires a cluster of machines and a different class of algorithms to handle these data sources.

On the other hand, deep learning is a specific group of techniques in the branch of AI (artificial intelligence). It has enabled machines to become smarter, reaching almost human levels of accuracy with some basic tasks that they struggled to perform in the past decade, but now machines are starting to surpass even our levels of proficiency. For example, all improvements in face recognition, speech recognition (Siri, Alexa), etc., have been possible due to successful applications of deep learning.

Thanks to some recent advancements in both hardware and algorithms, these techniques are now having tremendous impacts in various fields of engineering. Transportation is no exception. We are surrounded by sensors everywhere, which are placed alongside roads and inside traffic signals, continuously sending out real-time traffic information. Even the mobile phones and other GPS (Global Positioning System) devices that we use in our everyday life for navigation purposes are acting as sensors disseminating traffic information. All this data, if used effectively, will help in providing smoother, safer travel for everyone.

How will big data and deep learning change transportation in the near future? How could they change transportation long term?

Big data has already started to impact the transportation industry. For example, the GPS devices that we use in our daily lives are used to get real-time traffic information, and this information is directed back to us, helping us know where there is traffic congestion, wait times to expect, and alternative routes. In the near future, these data sources and smart algorithms will help us build smarter cities and smarter transportation services where we do not need to wait in congested traffic, where we can reach our destinations quickly and safely, and where we can spend our precious time more productively. And long term, these improvements will ultimately lead to long-awaited autonomous cars becoming a reality, which will change our transportation industry entirely.

What are some of the big, current issues in transportation? Can the solution be found in big data and deep learning?

Our cities are growing bigger, not in terms of the size, but rather in terms of population. The number of cars on the road is increasing, but the roads are not expanding at the same pace. As a result, we are seeing growing congestion, longer waiting times, more accidents, and greater risks. Intelligent transportation systems can help in making our roads safer and smoother. Efficiently extracting real-time traffic information from different data sources can help in reducing congestion significantly by distributing traffic sensibly. Traffic information paired with weather information can help in providing users with accurate predictions of what to expect on the roads and even in adverse weather conditions like snow. Deep learning techniques can transform the cameras into human eyes, continuously monitoring traffic conditions, quickly detecting traffic accidents or adverse road conditions, and transferring that information to road users as quickly as possible, reducing their risks and thereby saving lives and time. Overall, big data and deep learning will not only play a major role in building smarter cities, but also in building safer rural transportation services too.

What do you think transportation will look like in the next 5 to 10 years? What big ideas will be coming into reality soon?

The transportation industry is expected to change a lot in the next 5 to 10 years. Although I am skeptical that autonomous cars will be in full force within the next 10 years, I do believe that connected vehicles will be the next generation of vehicles that will take over the market. In this system, cars can “talk” with each other and neighboring infrastructure, giving drivers a 360 degree overview of what’s happening around the vehicles. They will warn drivers regarding potential hazards—such as black ice, sudden braking of out-of-sight vehicles several cars ahead—and communicate with smart traffic lights, thereby reducing idle time, saving fuel, and reducing emissions.

As a graduate student getting ready to enter the transportation engineering field, what impact do you hope to make? What insights will you bring to your work?

Our focus is to develop systems that can be implemented on the fly in the real world and make useful impacts to society. My present work is a small part of that broad scope where I am trying to detect traffic incidents quickly and accurately. However, my goal will always be to build systems that can make transportation systems safer and life easier for all of us.

What advice do you have for youth hoping to enter the civil/transportation engineering field?

Remember, you are entering the field at a very exciting time when things are changing rapidly. New technologies are entering the market and leaving a strong mark on the present as well as our potential future. Use your knowledge to extract as much information as you can from all the sources that you get. Transportation is a serious interdisciplinary field. So be prepared to learn things that you are not expecting. Learn with fun, “play” with the data, and the data will give you back rich, useful insights. As a wise man once said: “Knowledge is Power.” I would add an extra line to it: “Data gives you that knowledge.” Use the data effectively and you can make this world a much better place for every one of us.

Related links

Google Scholar page:

(Article) CCEE students sweep awards at regional, national Intelligent Transportation Systems contests:

ResearchGate page:

By Hannah Postlethwait, Go! Staff Writer

Go! Magazine Article Index