The Bullwhip Effect

My first reaction to reading the story about the continuing saga of San Fancisco’s new Muni line creating problems that are throwing the entire system into chaos (again mind you, this is a regular thing for Muni) was that for the $648 million the line extension cost to build one would think that someone could have created an effective model that better demonstrated the consequences this new safety device would cause:

At the intersection of Fourth and King streets, the T-Third line crosses in front of the J-Church line, creating a potential for train collisions. In an effort to prevent such accidents, engineers devised an automatic system that locks the T-line trains in place if an approaching J-Church train is at a boarding platform a block away.

But the system has created a bottleneck during the peak morning and afternoon commutes, backing up trains and causing service delays. The $648 million T-Third line started full service Monday, and problems rippled through the entire Muni Metro rail system.

Muni would have also done well to consult with non-transportation experts about something called the Bullwhip Effect in manufacturing supply chains. What is happening is that there is increased volatility in the system the further away from the observed problem and that very much is a symptom of poorly functioning systems as a whole. While transportation systems don’t increase/decrease supply dynamically, they do operate on a complex model that is built around forecasted demand and that is at the root of the Bullwhip Effect.

In reading the article it is also evident that there are at least 5 specific things that Muni has done which are amplifying the problems.

  1. The aformentioned "safety device" was well intentioned but ill-conceived
  2. A decision to not implement a signaling device that gives trains priority at one major intersection
  3. A poorly designed boarding platform that confuses riders and causes a potential pedestrian safety issue
  4. Rerouting the N-Judah train at the same time the new T-Third line went online
  5. Lastly and most significantly, Muni tested the new line for 3 months prior to the launch but only on weekends. The system never once received a weekday commute hour test before launching, a test that would have instantly revealed the shortcomings of the system.

Each of the 5 problems above has a human cause and is yet another example of a complex system that fails not because of a single problem but a series of problems that could, in this case, have been identified well before putting the system into production.

On a related note, Muni director Ford should be given high marks for his efforts during this crisis, which he inherited as a recently appointed director of the transportation system. He has been highly visible to the public, both in statemetns and media, and where it counts, riding the trains during commute hours and talking to the public and Muni employees about what they are doing to fix it. He appears to be demonstrating strong crisis management skills, but ultimately he will be judged on his ability to move beyond the immediate crisis and actually fix the system.

Tags: , , ,