|Jacki O'Neill, Stefania Castellani, Antonietta Grasso, Frédéric Roulland|
|Workshop "Ethnographies of diagnostic work" at Lancaster, UK, 17-18 April.|
This paper describes an ethnography of a call centre where troubleshooters attempt to diagnose and fix problems with large office devices, e.g. printers over the phone. The help-seeker is generally the user of the device, at the site the device is installed, and often non-expert in the technicalities of the device. Diagnosis and repair is a collaborative activity which consists of a number of activities. These include: 1) the collaborative working up of a problem description from customers’ initial reports; 2) translation from customers terminology to that of the device/troubleshooting resources and from the technical terminology of machine and knowledge-base to the customers language; 3) mediation – troubleshooters mediate between technology resources and customers and customers mediate between machines and troubleshooters. The distributed nature of the activity means that both troubleshooter and customer engage in co-ordination work to make this mediation work. The troubleshooter has access to knowledge whereas the customer has access to machine. Diagnosis and repair usually involves physical manipulations of machine or its software from instructions over the phone. The troubleshooter only has limited access to what the customer is doing and must verbalise all instructions. In addition there is often a dislocation between the site of the problem and the site of problem resolution, that is, telephones are rarely by shared office devices.
The ethnography emphasised the social nature of troubleshooting work and this understanding led to some inspirations for technology design to support such work along different dimensions: online support, collaborative troubleshooting, and between the customer and the machine. In this paper we first describe the case study findings and then we illustrate the technology ideas that they have inspired. The aim of the technology design was to bring critical features of the user-expert troubleshooting interaction into situations where troubleshooting might be undertaken without an expert; and where the expert remains to better support their work. Our wider aim is to enhance the design of the technical by basing it clearly in the social.