In recent years, contractor Try has built a new No 1 court and a 728-seat extension to the Centre Court as part of a masterplan by Building Design Partnership. This year, it handed over the Millennium Building, which adjoins the Centre Court and contains facilities for press, competitors and officials.
It is on this latest project that Try decided to pilot its zero-defects initiative, perhaps purred on by the competitive spirit of the place. “We identified Wimbledon because the deadlines are set in stone. Once the championships are on, our product is on display to everyone,” says commercial director John Homer. BDP was also a willing recruit to the war on defects. Project architect Roger Stollery says: “We had done something similar before, but never with the contractor’s assistance. The way the Try section managers have taken it on board has been the biggest benefit.”
The pilot, which is due to be rolled out on all Try’s jobs, first involved squeezing defects out of the construction phases and then achieving an early three-way agreement between client, architect and contractor on standards.
“If you put trades right as you go, you’re left mainly with aesthetics,” says David Bates, Try’s project co-ordinator.
However, eliminating trade blips is usually a gruelling process involving rounds of inspections followed by a marathon paper chase as subcontractors correct all the defects. Hence the birth of a computerised solution. As Bates says, “All that paperwork was just crying out for a database.”
So, Bates set out to devise a system that could be handled comfortably by IT-literate architects, section managers and “the guy on site with a paintbrush”. Bates’ system is based on Access, the standard database tool in the Microsoft Office suite. Try’s Bates turned down the chance to bring in a software specialist to design the defects database. “I wouldn’t have been able to modify it later without difficulty. I wanted it to be something we could all understand.” Instead, he invested in a copy of the jargon-free Access for Dummies handbook and devoted two months of last summer to creating the database.
The first task was to draw up a master index of predictable defects, created after Try managers filled in questionnaires on common errors. Fifteen broad bands were created, each allowing up to 1000 defects to be listed. These were called “thousand” categories. Examples of such bands included ceiling finishes, external joinery, floor finishes and concrete. These categories were then subdivided into groups that could each contain up to 100 defects.
For example, if ceiling finishes have a code of 12 000, defects to suspended ceilings might have a 12 400 classification. A damaged ceiling tile could then have a 12 401 reference, followed by 12 402 for a missing tile, 12 403 for a dirty one, and so on.
The system went live in January. Now, section managers inspect each room, equipped with a standard form, a photocopied plan of the area and a copy of the relevant part of the index. Working from the left of the door in a clockwise direction, and inspecting ceilings, walls and floors in a pre-agreed order, each defect is noted and its reference number entered on the pre-printed form. The location of each numbered defect can also be marked on the plan. “We’re trying to limit pencil-on-paper time,” says Bates.
The benefits of the Try system start to kick in back at the site office. Previously, managers would have typed up their handwritten notes, or had them typed, risking the introduction of errors. Then, the list would have been photocopied, and the defects highlighted in coloured markers to indicate work for colour-coded subcontractors. But at Wimbledon, the coded manual forms are put into templates on the PC by a member of the administrative staff. The system then instantly produces room-by-room, contractor-by-contractor print-outs of defects, with a full description of each problem extrapolated from the reference number, a location and a date by which the work should end.
“We can process the forms at a rate of knots, and they don’t get lost. But the biggest bonus is using the expertise on site efficiently – we’re not paying them to be clerks,” says Bates. He says that logging defects would otherwise have occupied 20% of the time of six site managers.
Try’s staff repeat the inspection process until satisfied that only a handful of defects remain. Then BDP staff start their rounds.
If an inspector encounters a defect that has not yet been assigned a reference number, the manager simply notes its description. Later, the system automatically assigns the next available number and updates the master file. Bates acknowledges that this could prove to be the database’s Achilles’ heel, because once the database is used more widely, the risk arises of different sites, or even two machines on a large site, assigning the same new number to separate defects.
But Bates does not view the system as rigid. “Why be a slave to procedure if individual site managers can make it work for them?” he asks. “Try is committed to it, but there’s still room for modification and experimentation.”