So you want to build a railway?

Jon McDonald, PE | 11 Nov 2014 | Comments

When I started in the rail business in the 1980s, technology consisted of telephone lines, signal control relays, and radios. Railway control was managed between central control and the train operator. Designing and building a rail line could be estimated and scheduled reliably by the foot or mile, because the approach to building railways had not changed much since the first signal system was built by William Robinson in 1872.

Since then, the rail systems industry has introduced a vast amount of new technology, as well as increasingly more complex methods of delivery—such as design-build and public-private partnerships—to help manage risk and lower costs. Meanwhile, communications technologies have transformed the transit environment. Yet rail projects are still planned, defined, and designed in much the same way as when I started. In fact, quite a few projects have roughly the same specifications, terms, and conditions as they did in the 1980s, before the Internet. It is unsurprising, then, that both “problem projects” and project costs have risen dramatically over the past few years.

The difficulty of bringing the rail industry into the 21st century comes not from the quality of engineering or technology that we use but from the philosophy of our approach. Seventeenth-century philosopher René Descartes theorized that the best way to study a complex system is to break it apart and study each part thoroughly. Today, we still follow this approach by separating each part of the railway into its respective components and assigning each component to a separate group of specialists to develop individual specifications. The contractor must then “reassemble” the parts, interpret what is missing, and incorporate the desires of operators and the public. Those gaps, differences of opinion, and newly realized needs result in delays and higher project costs.

Frustrated by the Descartes approach, German biologist Ludwig von Bertalanffy developed a different approach in the late 1960s, proposing that an organism should be studied in whole as a complete “system” rather than in parts. A system, as defined by Bertalanffy and subsequently the International Council on Systems Engineering (INCOSE), is a combination of interacting elements organized to achieve one or more stated purposes. Systems engineering is then an interdisciplinary approach to designing successful systems, using a process that takes into account all components and stakeholders to describe a system’s attributes over their entire life cycle.

When engineers apply a systems approach to rail, the railway is viewed as a single system that delivers passengers to and from their destinations. This system exists within an ecosystem of passengers, cities, regions, and geographies, and may have attributes like “safe, secure, and reliable” because the stakeholders within the ecosystem needed or wanted those system attributes.

INCOSE has developed a systems engineering process flow to help decision-makers determine not only how to design, build, and procure the right thing but also how to design, build, and procure it in the right way. This process supports a full life cycle approach to each project through activities such as:

  • Defining, evaluating, and documenting stakeholder and system requirements up front.
  • Exploring potential solutions—discovering new ideas, solving potential issues, and determining the market’s capacity to deliver exactly what has been specified.
  • Designing the final system based on the agreed-upon goals and needs, building visibility and checks and balances into the process.
  • Verifying and validating the solutions to ensure long-term project success.

Life cycle costs should be considered, not just the initial capital cost of the rail system. For example, statistics show that the initial cost of purchasing a Boeing 737 aircraft represents just 5 percent of its total whole-life cost. Similar issues arise in rail, and designing with life cycle costs in mind can save the agency and the public millions of long-term residual dollars spent.

Systems engineering has had a number of successes, including the London Underground’s Jubilee Line upgrade, which failed on two other attempts to implement communications-based train control. The UK’s West Coast Mainline project was also experiencing difficulty before systems engineering was introduced, and the Mecca Metro in Saudi Arabia was rescued by systems engineering mid-project. In each of these cases, the failure was trying to build an isolated project without considering how it fits within the ecosystem of the public, organization, and existing system.

So when thinking about building a new rail line or extension, it’s important to consider how significantly railways and public transit have changed over the past few decades. The equipment and technology we use have changed, as have our customers’ demands and the overall transit experience.

Systems engineering has been used to deliver high-profile projects successfully, pleasing both the agencies responsible for them and the communities they serve. To succeed in today’s environment, a new approach is vital—one that considers the “system” rather than a sum of individual parts.