Managing rail disruptions the easy way

Busy subway station

If you’re chuckling at that heading, I don’t blame you. Managing rail disruptions is far from easy.

But here’s the thing: there is an easy solution. Easy, mind you – not simple. There are no simple solutions to rail disruptions, and if I leave you with one thought in this post, it’s this: Beware of simple solutions to monumentally complex problems.

How complex is a rail disruption anyway?
We all know that rail disruptions are incredibly difficult to resolve swiftly. Within minutes, planners have to create a temporary schedule that:

  • Minimizes inconvenience to passengers.
  • Minimizes the recovery period in which the temporary schedule is in operation, so that services return to normal as soon as possible.
  • Minimizes operating costs (all other things being equal).

What do planners need to take into account?

As it turns out, a lot.

Among other things, planners have to consider:

  • Expected passenger flows: Where are the passengers and where are they heading?
  • The position of rolling stock: What’s the best way to deploy rolling stock to address those passenger flows, while avoiding dead-end trips?
  • Working time directives that affect crew: Where are the crew? Can we reposition them to ensure that the relevant people are where they’re needed at the right time? Have we allocated tasks to avoid over and underutilization?

The time pressure is tremendous. Planners are well aware that as they fumble towards a solution, the disruption is snowballing into a major incident.

Why keeping things simple is never the answer
The simple answer is because you can’t. You just can’t.

On top of the complexity, there are high – and unavoidable – levels of uncertainty. For example, the disruption could well take longer than anticipated, with knock-on effects on the revised crew and rolling stock assignments. This means that planners are also struggling to create robust schedules that factor in uncertainty and provide them with the flexibility to react to further developments. Without building in some slack, each ‘solution’ risks creating further disruptions.

In spite of the complexity and uncertainty of rail disruptions, many rail companies are still trying to get by with simple solutions.

Here are some of the usual suspects.

(1) Build in buffers
It’s possible to prepare for disruptions by deploying more resources than are actually needed. For example, you might employ additional crew, and place extra rolling stock over your network.

Verdict: An inefficient and costly solution. For most network operators, the density of their rail network rules it out.

2) Resolve disruptions manually
The mind-boggling array of scenarios that need to be created and evaluated is akin to high-level chess. Human planners simply aren’t capable of quickly and continuously thinking at least six moves ahead.

Verdict: Not a realistic option – unless you’re willing to tolerate long recovery times and high costs.

3) Create set responses
This assumes that it’s possible to predefine various scenarios and call up a set response when a particular situation occurs.

Chess comes to mind again. This ‘solution’ is rather like hoping to win a chess game with preplanned scenarios for every possible situation on the board. There are zillions of possible scenarios. Where’s the rolling stock? The crew? What exactly is going on in terms of track availability and passenger flows? How will business rules and working time directives affect possible solutions?

Verdict: Good luck with that. If only life – and rail disruptions – were that simple.

The easy way out – and yes, there is one
Instead of trying to create set responses for zillions of situations, intelligent planning and optimization takes on board your business rules and constraints, and uses smart algorithms to generate possible solutions. These real-time scenarios reflect what is actually happening on the ground. Solutions can then be ranked according to how well they satisfy your KPIs.

While control systems provide visibility into the chaos on the ground, resolving that chaos requires something far more sophisticated. Over the last decade, huge progress has been made in developing the advanced optimization techniques that are required to master rail disruptions. To discover more about what intelligent real-time planning and optimization has to offer rail companies, here’s a quick read on a complex subject: Taming Passenger Rail Disruptions.