Paint Flow Project Version #5, Still Underway

Me with wife Heide's painting in the background

The possible collaboration of UEI and Vivid Inc for the new Paint Flow Control (PFC) Project is still underway with Thomas Nguyen traveling to China to visit a painting company which is painting keypads and contact surfaces for remote control units for UEI. Thomas was enlisted to go to the company to learn about their painting equipment and process and to assess their system with the aim to make recommendations for improving their process and system. Thomas spent only a couple of days there. He met with a UEI engineer who escorted Thomas to the painting company. Thomas observed the company's painting process and made suggestions for improving the yield of the process as their reject rate is quite high at 30%. However, the company has no incentive to improve the acceptance rate of their product as UEI pays for all their output, good and bad.

When Thomas returned from China, we learned that rather than manufacture my control board design, UEI wants to use a commercial control board for the new project. He wants to use the Digi International BL2600 that I mentioned in previous posts and had used for our ill-fated Silk Screen (SS) Project. I like the idea of using the BL2600 as it is plenty fast enough, has analog I/O, and has the terrific Dynamic-C compiler that was originally developed by Z-World in Davis, California many years ago. Z-World became Rabbit Semiconductor and then Digi International acquired Rabbit Semiconductor in 2005, continuing the Rabbit Semiconductor product line. A special feature of Dynamic-C is a real-time-operating system, so it's ideal for control systems.

I had done a lot of software development for the BL2600 on the SS Project, reported previously, and so have a head-start for the new PFC Project. The BL2600 also has an Ethernet interface that allows communication from a PC to set control parameters and acquire process data. I used National Instruments LabWindows/CVI software for the SS Project and plan to use LW/CVI for this incarnation of the PFC system. I had programmed a Fuzzy Logic algorithm into the SS control and plan to use FL for this new PFC project. However, the PFC system may have different dynamics than the SS project as the SS system was extremely slow and the painting systems are considerably faster. So I will have to do some analysis to come up with a new FL algorithm specifically tailored to the new PFC system. Also the SS system utilized three ink pumps, whereas the latest PFC system may not require a pump as Thomas hopes to be able to use a gravity feed system to provide adequate pressure. Although I'm very respectful of Thomas' experience and expertise for painting control systems, I confess to being skeptical of his gravity feed idea as previous painting projects always used a pump and I suspect gravity feed won't provide adequate pressure for the paint gun. Exactly how the system is to be configured is to be seen.

So I figure that the system design still up in the air but I have started working on a schedule and cost estimate that I will be able to adjust as necessary. Once the system details are ironed out, I will update my schedule and cost estimate, then I plan to negotiate a contract for the system development. Hopefully all the system details are worked out soon and we can work out a deal for my development work so I can go ahead with the project. Just today I learned from Thomas that Vivid Inc plans to obtain the robot and other equipment from Chuck Nelson, equipment that includes one of my controller boards and other equipment that Chuck had put together for the Control Coat project that I previously reported on. Thomas expects that Vivid's acquisition of the robot and other equipment from Chuck will simplify the overall system. So this is encouraging news, indicating that the project is going ahead, albeit haltingly.