"It will be a journey, but if each step of the journey creates incremental value for railroads, they are going to enjoy it."
Today SmartRail World visits the offices of Optym, who since their founding in 2000 have created optimization, simulation and data analytics solutions for Fortune 500 companies around the world in the railroad, airline, trucking, mining, shipping and retail industries. And it’s their world in the railroad sector where clients include CSX Transportation, BNSF and Canadian Pacific Railway that we are focusing on today as we speak to the founder of Optym, Dr. Ravindra K. Ahuja. Dr Ahuja was a former Professor at the Massachusetts Institute of Technology (MIT) before embarking on his business career and offers some fascinating insights into the future of our industry.
Luke Upton (LU): Hello Ravindra, Optym works with some of the world′s largest railroad carriers – but for our readers who are unaware of your work, perhaps you could offer a short introduction to your work in the rail sector?
Ravindra K. Ahuja (RKA): Hello Luke, decision problems faced by railroads are very complex and Optym is performing cutting-edge research on these problems and bringing the fruits of its research to railroads. Optym was founded by me (an established academician) and employs people with significant research backgrounds. We have a very strong focus on research and development, having built yard simulation systems to create yard layouts with increased throughput capacity. We are now building real-time yard management systems. Optym has produced rail network simulation systems to determine network congestion points and the best investments to increase capacity. We have also created a real-time train movement planning system to direct the movement of trains in the network. Our products also include data visualization and analytics systems to measure the health of the rail network and identify operational hotspots. In the future, Optym aspires to be the company that will drive the movements of all assets (trains, railcars, locomotives, crews) for railroads and create unparalleled efficiency for them.
LU: The rail industry is an ever more complex network of tracks, yards, trains, locomotives and crews. How can the challenge of the planning and scheduling of these assets effectively be overcome?
RKA: Indeed, the rail industry is comprised of complex networks of tracks, yards, trains, locomotives and crews. The complexity of these networks is far beyond a human brain to fathom. Despite this, planning and scheduling of its assets is done manually with little machine intelligence due to a variety of barriers. But now the stars are lining up and a rail renaissance is on the way as the main barriers are being lifted:
Data Availability and Quality: Railroads do not have data available for all of their assets, and the quality of what is available is questionable. Railroads need to invest in ensuring that they have good quality data available for real-time decision making and planning. With pervasiveness of GPS technology and the mandatory requirement of the PTC technology, high quality data is becoming available in real-time and this will alleviate real-time data related concerns.
Investment in Soft Technologies: Railroads do not hesitate in making big investments in hard assets such as locomotives and railcars, but they are quite reluctant to invest in software. They need to realize that better software can reduce asset costs and can offer benefits in cost and service. They need to realize that the future railroads will be digital railroads and machine intelligence will drive operations. Leadership at several railroads has started realizing the need of intelligent software and we are witnessing renewed interest.
Community Supported Model: Traditionally, railroads have been building home-grown systems and have not benefited much from the community supported model. Railroads are now realizing that building and maintaining intelligent systems in-house is not financially viable. They need to open their doors to qualified vendors and partner with them to build such systems that can be commercialized and supported by several railroads instead of one railroad.
Algorithmic and Computational Advances: Railroad decision problems are very complex and difficult to model mathematically. The high level of randomness in operations fuels the difficulty even further. We need fast algorithms, efficient programs and the latest hardware to solve these problems. Recent technological and computer advances have now made it possible to do so.
LU: How important is ‘real-time’ data for the railroads you work with?
RKA: Real-time data is very important for railroads. Let us illustrate through some examples - railroads need to know where their trains are at any time and whether they are maintaining safe distances. Crews need to know when trains that they are going to drive will be arriving. Maintenance crews performing track maintenances need to know when trains are coming to the track so that they can step aside. At yards where trains end their journey, managers need to know when inbound trains are arriving so that they can reassign locomotives to outbound trains. Good quality and reliable real-time data is central to operating a railroad efficiently.
LU: … and what kind of data are they demanding?
RKA: They are demanding data for all of their assets: trains, locomotives, railcars, crews, and tracks. Where they are at any time, where they are going to be in future, their health, whether they can be used and for how long. Just as you cannot cook a meal without ingredients, you cannot build a real-time optimization system without real-time data feeds.
(See also: Big Data and the rail industry: An expert guide.)
LU: And this message resonates well with railroads?
RKS: Yes, but for railroads to invest in software you have to provide a very strong product. Railroads don’t typically like investing in software – sometimes a suspicion of buying something they can’t see. Although this is now changing. We often hear railroads, only half-jokingly, say they want to be the “second client” for any product, hoping to minimise any risk they feel they might be exposed to with a new tool. So this is where our extensive research and development come into play in both proving the product prior to purchase and then detailing the successful results afterwards. You have to work hard with railroads!
LU: With this in mind, rail planning systems have changed considerably in recent years and continue to evolve. What will the next generation offer?
RKA: Rail planning systems have not evolved much in the past two decades and most planning is still manual and uses experience and gut feeling. The need, however, continues to grow and the tipping point has been reached. Next generation systems are now being developed and will gradually start changing the landscape. They will offer data visualization system to see the current status of all assets, enabling a deep-dive analysis of what happened. They will provide planning systems so that planners can create future scenarios and analyse them. They will provide real-time management systems that will govern the efficient movement of all assets, ensuring they are productive and creating value for railroads. It will be a journey, but if each step of the journey creates incremental value for railroads, they are going to enjoy it.
LU: And finally, you founded Optym after a lengthy career in teaching at several prestigious universities in the US and India. Do you miss teaching?
RKA: Ha! Well, I loved working in universities and I’m still a Professor at the University of Florida, Gainesville and although I don’t teach in a lecture hall anymore, I do still feel like a teacher except now I’m sharing knowledge with railroad personnel and executives instead of students. And the nature of our changing industry and the developments that we are seeing means that I’m still learning myself every day. Teaching career is about constant learning and disseminating your knowledge to others and this is still happening every day.
LU: Many thanks Ravi! Looking forward to hearing more again soon.