The CIS case study was part of a long-term research cooperation in the areas of HCI and CSCW involving Danfoss, the Computer Science Department at Aarhus Univer-sity and the Art, Culture and Communication department at Malmö UniverUniver-sity College, and four wastewater treatment plants in Denmark and Sweden. Methodologically, the project was coordinated across research groups even though, thematically, the three research groups worked from different research goals or themes. Thus, field studies of three wastewater treatment plants were the basis for initial investigations and familiari-sation with the wastewater treatment area for researchers from all three participating organisations. Later, feedback sessions and design and evaluation workshops with and without users were the primary way of sharing and comparing findings with the other research groups and users. For ’my’ group, the purpose was to explore the theoretical notion and practical design of common information spaces, i.e. how information is collected, shared and accessed in a wastewater treatment environment.
We spent 40–50 hours within a five-month period doing participant observation at the site, following workers through their entire daily routine. Different researchers fol-lowed different workers, using hand-held video cameras or a digital camera to capture the events. We analysed the video with special emphasis on the daily work practice, use of artifacts, and how workers dealt with the disruptions, using the techniques described in [Bødker, 1996] and [Buur and Søndergaard, 2000]. Video clips were presented for the plant employees at three feedback sessions/user workshops at the plant. Further-more, we compared the work practices at this site with data collected by other project researchers at the other wastewater treatment plants and held 3 project workshops em-phasising design and evaluation with participation from all the cooperating parties.
47
7.1.1 The plant
To establish the area of scrutiny to the reader, we shall briefly describe the specific type of wastewater plant we have visited, and look at how the work is organised. The wastewater treatment plant (MR) was one of the first larger plants in Denmark to imple-ment automatic process optimisation for the removal of nitrogen. The automation has been possible due to the development of new technology, sensors, which has allowed for on-line measuring and control of the primary parameters of operation. Not surpris-ingly, the process optimisation has radically decreased the expense of e.g. the use of chemicals and in relation to the airing process. The plant has an estimated capacity of 220.000 Personal Units (PU) but is constantly running at 110–150% because the plant has not been able to expand to match the increase in the city’s production of wastewa-ter. The purification process includes both mechanical, chemical and biological phases and the segmented sludge is first used to produce gas in putrefication tanks for running the plant before it is pressed and taken away to an incinerator plant.
7.1.2 The work
The studied wastewater treatment plant employs 8 people, 2 of which represent the management. The area of responsibility for the management is the plant in general whereas the workers are divided into sub-groups with more specific areas of responsi-bility, as we shall see in the following. Also, the management have goals and interests that may be in conflict with those of the workers. Among the workers we have the following divisions:
• 2 workers responsible for the lab, receiving the sludge-trucks, area of preliminary sorting and sand trap.
• 2 workers covering most of the outside areas of the plant as well as the putrefi-cation tanks, sludge tanks, gas-turbine building and the control room
• 2 workers working in the building with the sludge press
The division of labour can be described as process-defined in the sense that the three groups of workers each are responsible for the tasks associated with a specific part of the treatment process. The division of labour is quite clearly defined and re-arranging of tasks only very rarely happen on a permanent basis but will occur temporarily e.g. in case of illness where the most pressing tasks are distributed among the rest of the work force. This organisation of work means that tasks cannot be interchanged between all workers but that each worker has a relatively strictly defined set of work tasks, with some overlapping. However, within these bounds each worker is able to ’juggle’ the tasks as the situation demands. In general, the work day has a very dynamic structure—
profound understanding for ones assignments and the plant in general makes it possible to e.g. place equipment in the area for later use, and to redefine the order of tasks in order to cope with the numerous events that cannot be anticipated. Furthermore, the workers give special attention to things that may later on obstruct the running of the plant (e.g. a motor sounding strange or leaking oil). To get a better understanding of the extend to which unplanned incidents occur, we might consider which events that may affect the work routines:
• Machine failure/alarms
7.1. THE CIS CASE 49
• Signs of problems
• Experiments on the plant
• Guests
The large number of alarms in one day alone makes it obvious that a great deal of flexibility is necessary in the daily routines to be able to keep the plant running smoothly. If the alarm is due to machine failure, it is often necessary for several of the workers to completely break off their daily routine and instead enter into a cooperative effort they are not usually part of. If the alarm is merely a warning of a breakdown, or if a worker sees signs of problems like a motor leaking oil or sounding strange, it is entirely up to the situation whether or not this calls for immediate action or can wait for days, whether the observant worker can handle the situation on his own or if other specialists need to be summoned. Noticing specific warning alarms and the physical signs of problems are highly prioritised in the everyday work as it naturally is much more desirable to prevent machinery breakdown than to recover from them.
Furthermore, at this particular wastewater plant, many experiments to optimise the pu-rification process in terms of cleaner water and decreased expenses are continuously running. Experimentation usually involves introducing new technology or work prac-tices, and is as such a high disturbance factor as these often provoke unanticipated events or effects on everyday work. Finally, the wastewater treatment plant very often visited by outside guests, especially school outings, and even though these visits are carefully planned and executed, they nonetheless present a disturbance in the overall running of the plant. Considering what a multitude of unplanned events on several dif-ferent levels these factors combined may cause makes it obvious that a high degree of flexibility in the work practices is crucial for the workers to maintain the continuous operation of the plant.
7.1.3 The existing web-of-technology
Looking at the current situation at the plant with regards to the technology they use, the predominant element is without doubt the central control system which maintains and controls the purification process and is accessed from the main control room or the manager’s office. Certain changes to the lower levels of the control system can, however, only be done from a console in the central control room. Furthermore, access to parts of the control system or sub-systems is available through two computers in local control rooms at the plant, one in the preliminary sorting area and one in the building with the sludge press. Other technological artifacts used in the daily work in the lab are the spectrometer which analyses the contents of the water in the samples the lab technician collects on his daily round connected to a matrix printer which prints out the results. The lab worker brings this print-out to the manager’s office or the central control room to enter the analysis results into the central control system because he does not have access to the central control system from the laboratory. While a computer in the lab is desirable for the lab worker, it is not considered important enough to be economically possible to the manager. This ’island’ is vital to the control of the water purification process because the lab is the only part of the process where detailed information about the contents of the water can be given. However, direct connection to the control system would not alter the fact that the laboratory information provides an average picture of the state of the water over a 24-hour period prior to the analysis, which is the consequence of the method of sample gathering as it is done today. Finally,
telephones and mobiles phones are used on the plant, but interestingly enough they are primarily used for communication with people outside of the plant or to call a co-worker on the plant from a meeting or other activity outside of the plant. Between the workers on the plant, telephones are very rarely used; even the unplanned events are rarely time-critical so information is propagated through morning meetings or at lunch, or the wastewater operators ’visit’ each other in their respective parts of the plant. Knowing each other’s routines, the wastewater operators have a very keen sense of where to find each other during the day. A well-functioning social system like this would be difficult and even pointless to try to support through new technology whereas I see possibilities in connecting the ’island’ to the rest of the system and generally provide a better connection to the components and the process outside of the control room.