Lots of progress in the last couple of weeks! I met with Phil at the glass show. He was kind enough to offer to send out one of his technicians to help us finish our implementation. So I went back to the group and had them generate a list of all items we need from SoftTech. This includes bugs, minor annoyances, and enhancements. There are only 11 items on our list. This week SoftTech killed 5 of them. By this time next week we may be down to just the enhancements.I pulled the whole group together and we went back through our workflow. With the input from the group, we have been able to document most of our process and simplify it down to some pretty basic steps. We have since implemented this on two projects and they are having significant gains in productivity. The last workflow issue to resolve is with BIM projects. The BIM models are required early in the process (around the time the shop drawings are being submitted for the first time). On a custom system project, it requires that you import all of the extrusions and create all of the assemblies before you can build the elevations and then export the model. The difficulty is with the milling ops. It is not easy to anticipate all of the required milling ops if you are working from the estimating takeoff or architectural drawings. If you wait to add the milling ops, then you may have to go back to all of your elevations and substitute the assemblies to get them to show up. We’re still working on how to address this.

Here are some of the lessons I’ve learned over the last couple of years working in V6:1. Be sure you have the right people. We’re in the construction industry. Do you have the technical people to do the implementation? You need people who like to customize software. People who are already writing Autocad code or excel macros. People who understand what is possible.
2. Allow enough time. I allowed 1 year to fully understand how we wanted to implement the software. It took 2 years. You can probably shorten this time period by hiring someone who knows your business to help with the implementation. At the time, no one was using the program for what we were doing. There was a huge learning curve.
3. Start small. Work on one department at a time. You will have issues and it is harder to get a lot of people to buy in when things go wrong. Your technical people will understand you are in beta and stay the course, but less technical people will have no patience for software with bugs. And they will be vocal. And you’ll have your boss or peers asking you lots of questions. Not everyone wants you to succeed. Sorry, it’s true.

Heading off to Glassbuild in Las Vegas this week. Look forward to seeing the folks from SoftTech as well as Emmegi. I’ll be on the floor Wednesday all day. Anyone else planning on going?If you are in Las Vegas for the show, stop by and check out the Harmon Hotel at the City Center. It’s the oval shaped building right on the Las Vegas strip. Our Las Vegas office just recently completed this project. Beautiful blue glass. You can’t miss it.
I was turned on to some books on Lean a couple of years ago and have been studying how it might be applied in our company. We just completed a book club on “The Toyota Way” with our Contractor on the Hospital out in San Francisco. Sutter Health has been sponsoring the classes. I signed up for a class at Stanford on Plan, Do, Check, Act. Classes start this month. I’ll post my progress.
Based on the Lean concepts, we’ve started some testing using smaller batches for our shop releases. V6 allows us to slice and dice our releases to match the customers needs in the field. This has reduced the amount of inventory and the time it takes to find parts for assembly for our unitized wall systems.

Haven’t posted in a month. Where does the time go? We’ve had an interesting August. If we were in Italy, I’d be writing about how well all of our vacations went (they get the whole month off). Instead, I’ll write about our V6 implementation.Ran into some snags. First, we were trying to use the elevations from V6 in our shop drawings. While this is entirely possible, the problem we ran into was with revisions. The draftsmen were taking the V6 elevations and modifying them to make them look more presentable. Changing hatch patterns, adding panels around the perimeter, datums, etc… Then every time we would have to re-output the elevations (to get key sheets, dimensional changes, etc…) we had to re-do the shop drawing outputs. This was taking hours. We have a really customized Autocad setup and drawing elevations is mostly done through a dialog box (much like V6) and this was never really a problem anyway so we’ve backed off of the shop drawing implementation for a while. We’ll get there, but too much too soon.
Second problem was we were getting some crazy errors and couldn’t save our quotes. SoftTech chased those down to a bad formula in a Bag (our fault) but it was frustrating the users to no end. We’re at that really delicate point in the implementation where we’ve completely switched from the custom software we were using to V6 and most of the users are not used to the new process. Because of the errors and the change in process they are feeling like things were easier the old way.
So we met yesterday to review the actual workflow. I think most of the problem is that we are trying to duplicate our old process with the new software. There are some amazing things you can do in V6 that simplify the checking process that we could never have done in the software we were using (I know because I wrote it). I think after the meeting that everyone realized how much simpler our process can be.
The last thing was that we had some bugs in the wireframe import. This is a new feature added to V6 and we are really putting it to the test on the LAX project. Very complex geometry with a unitized wall system. SoftTech thought they had it fixed in their last release but it wasn’t and it took another 2 weeks to deliver the changes. We got them yesterday and tested the release and everything is now working.
Once you get past the software implementation you have to work through the people component. If the software is a huge improvement on your existing process, then this will be easy. If it is only incremental, then it is very difficult. V6 can do about 80% of what our old software was able to do. It can also do other things (like BIM outputs) that our old software could never do. V6 also builds on itself and that will be the biggest gain. Every new project has setup and checking but every subsequent project with that system will be a walk in the park. Our old software was at the end of it’s ability to improve. We’ve only scratched the surface of what this software can do. But for the guy trying to finish his fabrication drawings for the shop by Monday, none of that matters.
So, August was a tough month. All projects are now going through V6 and if we can get a few glitch free weeks and improve the workflow by using the V6 features I think we’ll have a good September.

We have our last round of training for Fremont this week. After this week, all of the engineering group in Fremont will have completed the V6 training. We are putting V6 through its paces right now. The LAX project (concourse and core) will be an excellent test of how well V6 will work for us. Complex geometry, BIM requirements, extremely fast schedule and totally custom system. So far so good! We are finishing the mock-up this week and have created wireframe geometry for the North Concourse.
We are also finishing the material takeoffs and starting fabrication tickets for the Samaritan office building in San Jose. Nice unitized wall system project. In fact, all new projects will go through V6. We are very close to finishing our implementation. Just a few things to finish on details (probably next week).

Spent this week training 4 more people on V6. The 27 or so training videos we have put together have made the training process much better and more professional. It also documents our process really well as well as providing an easy way to go back and review specific items if you can’t remember something.
So we are very close to completing our implementation. Really just a few items left to go. We need to complete dialing in our data access from other branches, finish our detail implementation for shop drawings and a couple of open items on packaging and pricing for miscellaneous items.
We are now using V6 on all upcoming projects. The Samaritan project, LAX, Neuroscience, Equinix and Kaiser Redwood City are our first projects for V6. Should have some results to post in the next month or so…