My Robots

Auto-Place - AP-C2

The first robot I can take some credit for is AP-C2 while working at Auto-Place. This was a robot that could roll dice onto a table, locate their position and orientation and pick them up. I received a patent for it.

APC2_2

Auto-Place had a product line of pneumatic pick-and-place robots. I used a Series 50 robot with a DC servo X-Y table and came up with a stepper driven rotating gripper and rotating camera. An overhead camera was able to roughly locate the dice on the table. The X-Y table drove the robot to the correct location where the arm mounted camera could see the dice with much more clarity, fine position the robot and determine it's orientation. It was amazing to watch it work. Overall the reliability was a big problem. It was controlled by and Imsai Z80 computer that had some real issues with electrical noise. But it did work. Unfortunately I haven't been able to locate any video of it in operation. Camcorders were not around back then. Later we converted the controls over to use a Digital Equipment Comp. LSI 11/3. We re-wrote all of the assembly language code and the main program was written in Fortran.

Copperweld Robotics - Souped Up CR-10

After Copperweld Marketing people decided that we need heavy fiberglass covers on the robots to improve sales, we quickly noticed that the performance of the machines was suffering with all the new weight. I designed a new robot base that replaced the polygon piston with a hollow rotary actuator with THK ball spline nuts. I then attached the THK shaft to a pancake cylinder underneath for the lift motion. The results was much more than I anticipated. The new body could rotate the arm, covers and all, 180 degrees in 1/2 second. The problem now was stopping the arm and keeping everything else attached. I have a healthy respect for robot safety and never turned my back on this robot. Ken Hornacek and I also worked a lot with on the seals, surface finish and lubrication in an attempt to get the seals to last as long as an 'O' ring.

CR10_1

CR10_2CR10_3

Unfortunately, by then our Marketing Department was convinced there was no market for this type of robot, even with the very stylish covers.

Graco Robotics EDY

I was hired by Norm Fender in June of 1990 to design a new low cost electric painting robot. GRI had been the dominant paint robot supplier for quite a while and I believe their biggest competitor in the US was Fanuc with the P150 electric paint robot. GRI had an impressive line of hydraulic robots and one electric machine, the EDX. The EDX was very complex and expensive because it was designed to be lead-through teachable. At the time, I didn't know anything about painting that didn't involve a brush, roller, or aerosol can. I worked for Dr. Bob Todd who lined me up on the requirements. Tom Wilson led the charge on electrical and explosion proof requirements (FM certification).

EDY_Paris_1

We completed the robot and had it running at the SITS finishing show in Paris by April, 1991. The above photo is the night before the show. (I'm looking for better photos) It ran great for the entire show. Unfortunately, Bob discovered a pinch point that got him an introduction to medical system in Paris . The pinch point was corrected quickly.

After the show, I organized a team from manufacturing and service to completely tear down and re-assemble the manipulator. I made some design improvements based on their feedback and ordered the parts for the second robot. About that time, GRI was sold to ABB. We had just sold several EDY's based on prototype demonstrations to a Korean automobile manufacturer to paint fascias. The direction was changed to supply ABB robots and the prototype was never assembled. I understand the guys in manufacturing bought a lot of beer with the scrap aluminum proceeds.

Going through some old notes, I came across my notes from the line-up given to me on the new "low cost servo robot" that was given to me by Dr. Bob Todd.

ABB - Fiat SEVEL (My Ferrari)

Shortly after the EDY was officially dead, my attention was focused on introducing the ABB painting robots built in Norway to the US market. As it turned out, GRI had quoted a very large order of robots to Fiat before the buyout. During the negotiation process, Fiat asked GRI for a quote for a dual arm robot to paint the inside of large cargo vans manufactured by SEVEL in a plant generally east and north of Rome. GRI didn't get the big order for robots, but ABB did get the order for the van painter after they bought out GRI. ABB refused the order more than once, but accepted after many discussions with Fiat purchasing. Jeff Demeter was chosen as project manager for a couple of days until I was informed by the newly appointed President, Nick Risvi, that he had decided that I would be the project manager and engineer on the project. The concept was a little better than a bar napkin sketch, but not much. It was December 23rd and the runoff was scheduled for early June. I had a lot of help from Julie Fischer and Lik Kwong on the CAD side. I rolled the plotter into my office and put in a lot of hours from Christmas through spring. Bob Todd, Joe Staub, Tom Wilson and Hank Kwong worked on the controls and we all got it done on time.

Sevel_1_800

This is the front half of the robot just before the runoff. Hydraulically powered with a 30Hp pump, the robot had a total of 13 axes and was run by two GRI HDX robot controllers. It had dual closed loop flow control and 24 color stacks with conventional guns. Each or the two guns would spray up to 1200cc/minute. Nearly 2 1/2 liters/minute is a lot of paint - what a fog inside a van! The robot would move into the vehicle at over 1 meter/second including the carriage and both robots.

Sevel_2_400Sevel_3_400

Above on the left is a photo of the robot partially inside the vehicle during programming. To the right is a rear view of the first robot on-site during commissioning, also partially inside the vehicle.

Sevel_4_400Sevel_NY_1_400

 

What made this challenging, besides the timeline and budget, was that the robot needed to paint with two applicators inside the confined space of a van. Even though they were pretty big vehicles, it would have taken an arm almost 4 meters long to reach all the way inside. Instead, I decided to use a standard ABB Trallfa hydraulic horizontal arms and wrists and put both of the entire arms inside the vehicle. This made for a very short vertical arm. My solution was to fix the vertical arm and use an articulating cantilever motion combined with the horizontal arms' motion to paint. This ended up reducing the main carriage movement to a handful of motions to zones where the arms could paint. The above left photo shows the arms' cantilever axes at different extension positions. Above right is a photo of Mr. Data, Mr Pieramattero, Myself, Bernard Choquel and Mr. Pittarelli being observed by NY's finest. The runoff was a success and we all went back to the plant wrap up details on the installation and interview contractors. Bernard was our sales guy at ABB who spoke fluent Italian, (just ask him). I never could have gotten it done without his help. Bernard also taught me how to drive in Italian. Did you know that 240 km/hr is almost 150mph? I've never fully recovered from Italian driving, but I do observe the speed limit now. I hired Didier Rouaud to do all the programming and installation supervision for the project. He had been laid off from his lab manager position at GRI France when ABB took over. After we got the first machine operational, Fiat ordered 3 more. I managed the construction of the last machines and Chad Quick took over as Project Manager working with Didier on the installation and commissioning.

More pictures, videos and observations later.

Content on this page requires a newer version of Adobe Flash Player.

Get Adobe Flash player

Back to Home Page