Railroads and the new math – big data and what it means

Commentary - Railways are in the midst of an upgrade. They probably won’t look very different from the trains we are used to seeing over the over the last few decades, but scratch the surface and their capabilities are moving fast into the 21st-century. Railroads are getting smarter and going digital. How?

The industry has a well-deserved reputation for embracing new technologies advances in order to improve their operations. Railroads were early adopter of RFID technology for data acquisition, equipping their rolling stock assets with Automatic Equipment Identification tags and deploying a limited number of AEI readers across their networks, along with other fixed, stationary sensors and data collection technologies, like hot box detection, acoustics sensors, wheel impact and hazmat detection, which have given the rail industry a treasure trove of data at their fingertips.

Evolving technologies – from wireless sensor networks to video analytics to wireless technologies and the potential of a continuous streaming of digital information – offer new opportunities to help railroads unlock the insights in their operations – through their data. And recent mandates, such as Positive Train Control, are accelerating this move to a digital railroad – with many implications and opportunities to be determined.

As the railroads delve deeper into the digital age, it is important that they continue to evolve their systems, capabilities and culture to take advantage of the data they collect to learn from and use it to better describe and predict real world events, as well as offer prescriptions for a smarter way of operating a railroad.

The railroad of the future will use analytics to mine the mountains of data, expanding their focus from thinking about operations in increments of months, weeks, and days to having information in hours, minutes and seconds. For example, operators could see patterns about the condition of their rolling stock, normally hidden in the data, that could indicate potential problems, giving them the opportunity to resolve them before they become expensive service delays. The potential to understand and use the data while it is motion can help an operator optimize their network in real-time. Imagine how combining this with the immediacy of mobile technology in the field can have a significant impact on operations.

This kind of insight can be used to reduce yard congestion, dynamically alter schedules in the case of unanticipated delays or unusual events, offer prescriptive adjustments to train operation, and enable greater participation in the supply chain by having the information that shippers and other constituents want, when they need it. New services can take into account a variety of changing factors such as demand, weather, day of the week, capacity, etc. It’s a win-win for railroads and for their customers: The trains will move faster and endure fewer slow orders and derailments, and railroad operators can reduce maintenance costs and increase profits.

With railroads set to spend $12 billion on capital investments this year, consider how comparatively modest investment in innovation and technology can offer improvements in service quality, reduce expenses and increase profitability. Strategically using math and data analytics could help operators manage their rail networks more efficiently and add capacity without adding more costly physical infrastructure.

As demand for freight rail services increases – by current estimates it’s expected to surge 88% by 2035 – the industry must weigh how to invest in both physical and digital infrastructure to handle the future growth all while keeping an eye on maintaining their rail networks and growing earnings.

This also represents an opportunity for the growing passenger rail business in the US. Imagine the power of connecting and sharing data across multiple modes of passenger rail – high speed, intercity and metro/light rail – to provide passengers with seamless travel across their entire journey. By connecting and sharing data – schedules and routes, for example – between municipalities, transportation providers and travel booking websites, railroads and passengers would both benefit greatly. Railroads could see increased passenger loads and offer new services, creating additional revenue streams. Passengers could determine the most convenient, direct routes and book previously intricate itineraries wirelessly and quickly. Local governments would have a better understanding of traffic patterns and how to improve congestion.

IBM has worked with railroads for one hundred years. Through this successful partnership, we have helped develop some of the most significant technologies for the industry. But, there’s much more to be done.

In order to take advantage of the insights that can be found in all of this information, IT systems must evolve too. They are becoming more flexible and architected to consume data at the rate that business takes place – streaming in real-time. Just think about how the computing system behind Watson, the deep QA machine that beat Jeopardy! record holders Brad Rutter and Ken Jennings earlier this year, powered its ability to sift through oceans of data, mostly unstructured, in just a few seconds and arrive at the best answer.

Current projections suggest U.S. railroads could spend $148 billion repairing and upgrading existing lines over the next twenty-five years. Determining the right balance of investment is no easy task for rail executives, but where there is risk there is also enormous opportunity.

The companies that apply math and take advantage of the data in their business will not only access information in real-time and at the right time, but lead the way to becoming a truly digital railroad.

biography
Keith Dierkx is the director of the Global Rail Innovation Center and global rail segment leader at IBM, which celebrates its centennial anniversary in 2011.