An M.G. Locost Build

November 15, 2011

With nothing better to do at work, we set up our Locost build in Microsoft Project. I think we included every step, in a total of 18 major tasks. We used actual time for things we’ve already done, and estimated times for everything else. Assuming no more do-overs, which of course is crazy talk, the whole build should take just over 2100 hours. We’ve got just over 250 hours in it now, which means we're 12% done. However if the project grows to 3000 hours, which is practically inevitable, we're only 8% done. Actually, even that sounds pretty good.

Although we know things will take longer than expected, we can’t just add hours to the project plan indiscriminately by padding each task, because we can’t predict where the screw-ups will occur, only that they will occur. Kind of like quantum physics that way. If a task takes longer than expected, we’ll update the plan at that time, and that way we’ll have a complete record of how much time each step took. Measurement determines reality.

The Locost Build according to Microsoft

Microsoft Project lets you specify working times for each day of the week, and for any specific day, like Christmas or Superbowl Sunday, when we're unlikely to be working at all. We have it set for 24 hours a week, mostly Saturday and Sunday. As it happens we’ve averaged only about 20 hours a week so far, due to work, vacations, club events, race dates, parties, and other obligations. That will likely continue, but since we don’t know our schedule that far in advance, we can only take days away from the project as we learn about them.

With all of those caveats, we can still look at the schedule and see what we expect to be working on for the next year or so, at least in a perfect world, and also how much time things have taken so far. Of eighteen major tasks in the build, the longest is welding at 328 hours. For some reason the shortest turned out to be the steering setup, at only 36 hours, so maybe that's not a major task.

It’s interesting to note that the upper control arms took 94 hours, and the lower control arms—which are only 66% complete and could therefore take longer—are projected out at 116 hours. That 210 hour total is almost 10% of the entire build. No wonder they seemed to take forever. By the time the frame is built, we’ll be just under a thousand hours into the project. The Locost book says you’re about half done at this point, and up until now we really didn’t believe it. But a thousand hours. Turns out that’s actually pretty accurate.

Microsoft Project doesn’t only track your hours, it also calculates dates. According to the plan, welding will commence on Friday, February 24th, and wrap up on Friday May 11th. We start the engine transplant—and therefore need a donor—on July 28th. We’ll be working on the brakes in June, and the cooling system in September. Of course the most significant date of all is that magical day when all the work is done, and our Locost is presented to the DMV for certification and registration. That date is February 11th of 2014. Likely? Hardly, but you never know. We could get lucky and absolutely nothing else will go wrong for the rest of the build.


Our Locost
Our Build Plan
Building a Locost
Build Summary
Workshop Manual
Construction Manual
Non-Locost Stuff

Please Note: Our database is currently unavailable so you will not be able to browse through log entries. This happens sometimes, and it usually doesn't last long. We're sorry for the incovenience. Please try again later. Or in a few minutes.