Painting a Building: A Conceptual Reconstruction Based upon the IDEAL Model
Presentation by Jordan Vause
- General
- Stimulus for Change
- Establish Context
- Build Sponsorship
- Charter Infrastructure
- Characterize Current and Desired State
- Develop Recommendations
- Set Priorities
- Develop Approach
- Plan Actions
- Create Solution
- Test/Pilot Solution
- Refine Solution
- Install Solution
- Analyze and Validate
- Propose Future Actions
- Biography - Jordan Vause
- For More Information
General
The IDEAL Model v. 1.1, formulated by the Technology Adoption Architectures Team, describes a technology adoption life cycle consisting of five major phases. These phases are further divided into fourteen activities and a "quasi-activity". The purpose of the IDEAL model, as it was originally conceived, was to describe in general terms the stages through which an organization involved in software process improvement would pass. The new model was developed to be applicable not just to software process improvement, but to any improvement effort. This brief case study is an attempt to relate the new IDEAL model to an activity that is not software-related to illustrate how broadly it can be applied.
Stimulus for Change
The Stimulus for Change is the "quasi-activity" already mentioned. While not technically a part of any activity, it is the condition, event, or direction that indicates that some sort of change is needed. It therefore initiates an IDEAL cycle.
In this case study, the stimulus for action was the condition of the SEI building. The inside walls had not been painted for several years. As a result, they were dirty, smudged, cracked, scratched, and scraped to the point of being obvious to visitors and reflecting badly on the image of the SEI. In addition, the inside fittings -- the carpets, light fixtures, and furniture -- had become dirty, further contributing to this negative image and reinforcing the need for general repairs.
Establish Context
The obvious response to the stimulus for change was to clean the fittings and paint the walls. However, the SEI occupies a large building in which over 300 people work. The layout of the building resembles a labyrinth, with many hallways, offices, and small administrative spaces. The SEI staff is busy and some part of the staff is in the building at all times.
Individual offices and their contents, especially the computer equipment, are heavily used. Prolonged denial of access to these offices would significantly affect productivity. To clean and paint the building as quickly and economically as possible without disrupting normal SEI operations would be a major challenge.
Build Sponsorship
The sponsor in this case was the building's landlord, who had to agree to fund the services of the painting contractor. Every other organization involved came under the direct or indirect control of Jo Donatelli, the SEI Building Services Manager. Building sponsorship as we describe it here was the process of convincing the landlord that painting the building was a necessary thing to do.
Jo, as the "champion" of the improvement effort, could have gone about this in several ways. Perhaps the most orthodox way was for her to have persuaded the landlord (using the obvious condition of the building) that painting was needed. She chose, however, to invoke the terms of the original lease agreement, which stated that the building was to have been painted every five years -- by waiting ten years the landlord was in violation of the lease. This is a perfectly valid way to gain adequate sponsorship, though enthusiastic sponsorship may require different means. It all depends upon the nature of the effort.
Charter Infrastructure
The coordinator of the effort was Jo Donatelli; but she and her staff had to direct the efforts of at least six other organizations. Although each of these organizations existed independent of the painting effort, they formed a single temporary "meta-organization" for the purpose of the cleaning and painting project. This organization formed the infrastructure for the painting effort, with the duties of each component specified in advance.
The painting contractors, who worked for the landlord, were to handle all painting tasks. The cleaning crew, which worked for CMU, was to perform all scrubbing of furniture and carpets and cleaning of light fixtures. The telephone staff was to handle all telephone moves. The Computing Facilities (CF) staff was to ensure that all computer equipment was shut down as required and moved safely out of the way of painters and cleaners. Building Services (BS) staff was to ensure that all furniture and packed boxes were moved out of rooms and other spaces to be painted. The occupants of the offices were to ensure that all personal and work-related belongings were packed in boxes for moving.
Characterize Current and Desired State
The building in its current state had not been painted or fully cleaned in ten years. The walls and ceilings were dirty; the walls had cracks and holes in them; the paint was faded and flaking; and the light fixtures were dirty.
The desired state was a building in which all spaces had been scrubbed, all holes and cracks had been repaired, all lights had been cleaned, and all walls and ceilings had been repainted.
These statements may seem very similar to the stimulus for action, but they were in fact based upon a much more thorough assessment of the building during which particular problems were found and noted. The stimulus statement said simply that the building was dirty and needed painting; the current state of the building, as determined by inspection, showed for example that room 2310 had a deep crack in the far left corner, or that the wall across from the fourth floor mailboxes had a large gouge in it, or that a light cover in room 5514 was particularly filthy. The desired state, therefore, was not "a painted building," but a building in which all of the discrete flaws in the current state had been corrected.
Develop Recommendations
The recommendations were as follows:
- The SEI building should be cleaned and painted simultaneously through a collaboration of several groups.
- This effort be done during normal business operations and without any special time set aside.
- This effort should be done in cooperation with the occupants of the building rather than against them or in their absence.
Note that the recommendation was not "clean and paint the building." This was an obvious goal of the effort, determined in the "set context" activity and presented to the landlord in the "build sponsorship" activity. The recommendations here refer to the ways and means employed to get from the current state to the desired state, of which there were several. The entire staff of the SEI, for example, could have been ejected from the building. The building could have been painted first, then cleaned, or vice versa. A single contractor could have been used rather than the composite infrastructure already assembled.
One recommendation not made in this case, by the way, was to do nothing. Such a decision is always a valid option in the IDEAL model.
Set Priorities
Money and time were not the first priorities of this effort (although it will probably be the first priorities of many such efforts in the real world). The priorities, in order, were as follows:
- Health and convenience of the employees. Several SEI employees were allergic to paint fumes and could not be in the building when the paint was wet. Also, wet oil-based paint tends to damage clothing, so painting had to take place when building traffic was at a minimum.
- Normal SEI operations. Operations were to be disrupted as little as possible, even if this meant that painting had to take place at odd hours and in random room order. Painters would have to work around employee schedules, especially travel and vacation schedules. Computer equipment was to be left undisturbed as much as possible, even if this meant that painters and cleaners would have to work around equipment that was plugged in and running.
- Quality of the work. The quality of both the cleaning and the painting could not be compromised for any reason. However, quality could not take priority over safety and normal operations. It could and would take priority over timely completion.
- Timely completion. The effort had to be completed in four months.
Note that the priorities are not always derived from the recommendations. Some do and some don't; the need for timely completion is related directly to this effort, but the health and safety of the staff would be of high priority in any undertaking.
Develop Approach
The approach in the IDEAL model may seem similar to the recommendations, but this is not the case. The recommendations are based only upon the goals as stated in the "set context" activity and the current and desired states. The approach accounts for recommendations but expands or alters them as necessary to meet the priorities that have been set for the effort. For example, one recommendation was that everything take place during normal business hours. This recommendation is in accord with the second priority, but in conflict with the first priority since painting could not take place during normal business hours. An assumption made at the beginning was that computer equipment would be moved out of the rooms with the rest of the furniture. This assumption was changed to leave computer equipment in the rooms, turned on, so that owners would experience as little "down time" as possible.
The approach to be taken, based upon the goals of the effort, the recommendations, and the priorities, evolved to the following:
- All painting was to be done during off hours and all oil-based painting (which required two days drying time) was to be done over the weekends.
- Rooms were to be cleaned and painted over a period of three days each and the rooms were to be scheduled so that six rooms per day could be painted.
- Rooms were to be scheduled for cleaning and painting based as much as possible on employee preference. Employees were to sign up for time slots and were to pack their own belongings before leaving the office.
- Teams of furniture movers, CF staff, telephone staff, cleaners, and painters, were to enter and leave each room at set times over a three day period to perform their duties. This would allow each team to work continuously.
- Furniture and boxed items would be moved into the hallways, but computer equipment would be moved to the center of the room and left on.
- Communication between Jo Donatelli, the various groups she directed, and the staff of the SEI were to be maximized so as to increase synergy and reduce resistance.
Plan Actions
The plan for an improvement effort differs from both the recommendations and the approach both in degree and in substance. First, it takes the level of detail down to the point of "brass tacks." Everyone is told what to do, when to do it, how to do it, and so on. Second, it translates an approach that exists outside of time and space into a scheduled event that is tied to specific days and hours. The approach calls for each office to be completed in three days; the plan says which three days those are to be. The approach says that six rooms will be painted every night; the plan tells me when my room will be painted.
The plan for cleaning and painting the SEI building is too detailed to include here, but it was written to account for the goals as set out in the first IDEAL phase, the recommendations as set out in the second, and the approach as set out in the third. According to the plan, for example, six offices will be cleaned and painted during the three-day period of January 15-17. These offices are 3121, 3215, 3319, 5218, 5401, and 5403. In each case moving and cleaning will take place on January 15, painting on January 16, and moving on January 17. All six offices have been scheduled by their occupants for their own convenience. In one case, that of 3319, the occupant was on a three-day business trip during that time. On Friday, January 17, moving and cleaning will begin for the library. Since the library is a common area, the painting will take place over the weekend. Finally, during this week kitchens and entrance doors will be painted.
Create Solution
The solution can be described either as the procedure used to paint one room, or as the act of painting one room (painting the entire building is essentially an iterative solution). The IDEAL model allows either interpretation; for purposes of this study we will pursue the first.
Test/Pilot Solution
The first week of painting can be characterized as the "test/pilot solution" activity of the effort. Since it was based completely on the plan, it did not have the benefit of any previous testing, feedback, or inspection, and changes were made to the plan after the week's painting was complete.
Three major flaws in the painting procedure became clear during the first week:
- Coordination between SEI Computing Facilities, who had to move the computer equipment in each room, and Building Services, who could not clean the rooms until after this was done, broke down fairly quickly with some recrimination.
- The availability of boxes became a problem after it was discovered that people were requesting boxes weeks and months in advance of their painting dates and few were unpacking their boxes afterwards to pass them along to someone else.
- The move of furniture out of rooms and into the hallways became a fire-hazard. This had not been foreseen, but was nevertheless not a problem for most people. One employee did say that the situation was unacceptable for him.
Notice that this activity should be one of "verification" rather than of "validation." That is, the problems with the proposed solution are identified ("Are we painting properly?") but the wider issue of whether or not the solution itself is the right one ("Was painting a good idea?") is not addressed. In practice, the two often blend and one cannot continually ask the first question without also asking the second. The "analyze and validate" activity described below provides a formal opportunity for comparing the solution as implemented against the initial goals.
Refine Solution
The problems found in the "test/pilot" activity may or may not be corrected; the IDEAL model does not specify. For the SEI painting and cleaning project, the affect of the "test pilot" activity was as follows:
- The coordination problem was solved by having Computing Facilities move its equipment by 9:00 AM each day rather than 12:00 noon, and by eliminating any direct contact between Computing Facilities and Building Services (Jo Donatelli and her staff would serve as liaison if necessary).
- Jo purchased a large number of extra boxes so that anyone who wished could pack their office belongings early. The issue of employees keeping boxes after painting was not directly addressed (many employees have yet to unpack).
- The issue of fire safety was finessed slightly, though in the view of the SEI nobody was placed in increased danger. Any fire safety violation (such as furniture in a hallway) must be corrected within thirty days. Since the contents of any particular room were back in that room within forty-eight hours, the issue of fire safety was deemed not applicable. Egress for the employee who complained was never blocked.
Install Solution
The "install solution" activity in this case was the painting of the entire SEI building according to the revised procedure above.
This phase can be repeated more than once; multiple revisions of the solution can occur. But there is a difference between errors in the solution (which were identified and corrected in the last two activities) and errors in implementation. Many minor mistakes were detected during the cleaning and painting procedure. These were found in the course of regular inspections, shown to the workers responsible, and corrected immediately.
It was during this time that the last item listed above in the approach became very important. Since the cleaning and painting of the SEI building affected everyone, and since it was in spite of everything a disruptive operation, Jo Donatelli ensured that updates and status reports were issued continuously in the form of electronic mail, posted notices, and individual notifications for every room. Floor plans, showing each office as complete or incomplete, were maintained and updated continuously. Everyone on the staff knew, or should have known, the progress of the effort at all times. There was no formal feedback mechanism set up for this effort, however informal feedback was easily made through using e-mail, telephone calls, and personal contact.
Analyze and Validate
The cleaning and painting of the SEI building lasted from Thanksgiving 1996 to the end of March 1997. After the initial weeks, during which the "create solution," the "test/pilot solution," and the "refine solution" activities took place, the entire effort went smoothly.
When the last rooms had been painted, Jo Donatelli went through the building with the painters to inspect. The purpose of this activity in any effort is to compare the results of the improvement effort with its goals. In other words, to determine whether the original point of the exercise had been fulfilled. On the basis of her own inspection and of the feedback she received subsequently from SEI staff members, Jo decided that the goals had been met and that the effort had been a success -- more of a success than anyone had expected.
One purpose of this IDEAL activity is to collect and analyze the lessons learned from the effort; and of these there were several. During the cleaning and painting of the building, a parallel effort involving the move of SEI employees into the building from an outlying location was taking place. The two operations had not been planned with each other in mind, and as a result they tended to conflict rather than complement each other. The entire issue of boxes for packing could have been handled better, as could the issue of the computer equipment.
These lessons, and the analysis that resulted from them, have been documented by the SEI for use in subsequent efforts of a similar nature.
Propose Future Actions
The next major improvement in the SEI building will be the installation of new carpet. This will be an operation very similar to the cleaning and painting just described. For that reason, the same procedure, modified to take the lessons learned into account, can be used with only minor alterations.
Carpeting does have its own problems. Furniture, including computer equipment, will have to be moved out, but shelves (which do not touch the floor) will not have to be emptied; for this reason disruption to the employees will not be as serious. Night work will again be necessary. Carpet removal is, if anything, a more serious allergy threat than paint (due to dust, fibers, dust mites, etc.). carpeting or rolls of loose carpeting may also cause slips and falls. Finally, old carpet, unlike old paint, cannot be covered up. It will have to be removed from the building and discarded, which represents a significant addition to the process.
Biography - Jordan Vause
Jordan Vause is a former SEI Resident Affiliate (RA). He came to the SEI in June 1996 from Lockheed Martin Western Development Laboratories (WDL) in San Jose, California, where he had earlier worked for eleven years as a software engineer, a site support engineer, and a software engineering manager.
While at the SEI, Jordan was assigned to the core development team for the CMM® Integration Project. Jordan has also been working with the Technology Adoption Architectures Team, whose goal is to extend and recast the IDEAL model for process improvement into a tool for general technology transfer activities.
For More Information
For more information about the IDEAL model, contact customer-relations@sei.cmu.edu.