Response to Collaborative Activity and Technological Design: Task Coordination in the London Underground Control Rooms

Summary

This paper discusses how collaborative work is conducted in the control room of the London underground or subway system. The raw material of the paper is transcribed audio recordings and direct of observations in the control room. These observations focused mainly on two types of employees in the control room, Controllers and DIA's. The paper describes some recorded patterns of work during normal and abnormal activity. Since the DIA and Controller sit together at the same desk, it is noted that they work very closely together, but rarely directly communicate. Several instances are recorded about how a DIA will "overhear" a Controller's conversation and make an announcement even before the Controller is finished with his conversation. The paper also discussed the role and propagation of timetable information, which is recorded using plastic covered paper. In discussing the implications of this analysis, the paper suggests some tools that might assist the current environment. They suggest that a real-time display timetable might be an improvement over the plastic sheet solution. The authors also propose the use of simulations so that Controllers can determine the impact of a potential reformations (changes to the ordering or schedule).

Reflection

This paper does a very good job of describing the actual "process of work" in the control room. By recording actual conversations and observing behaviors they noted many detailed subtleties that appear to be involved in the work. This kind of analysis would be very useful when designing a control system. Two criticisms emerge from using this approach in a scientific paper. First, it would be interesting to know if this mode of working was common to all DIA's and Controllers, or just a few. The difficulty with the detailed analysis is that it may not be generalizable, even within the control room. Another fault is that the working patterns are treated as a given, and are not analyzed as part of the system. It might be possible that the working style of the controller and the DIA are a product of social conventions (due to rank, union rules, etc) or a drawback of their technology (no instant IM's, etc). It would be foolish to design a new system around a set of conventions that exist because of the limitations of an old system.