The Jefferson Project, Part 2: Automation as a Counterweight to Low Labor-Rate Assembly

Estimated reading time: 3 minutes

This is Part 2 of an interview I conducted at SMTAI with the founder of The Jefferson Project and the forthcoming Jefferson Institute of Technology, Tom Borkes. In Part 1, which ran in the January 2016 issue of SMT Magazine, Tom provided his well-researched plan to introduce students to tech manufacturing through a four-year, hands-on, real-world college learning experience that brings tech manufacturing to them. We also discussed the paper he presented at SMTAI, which focused on the concept of building meta-process control into an assembly operation’s infrastructure.

In Part 2, taken from the original interview, Tom expands on the example set forth in his SMTAI paper, and describes another important tool in reducing labor cost through the reduction of labor content: designing for automation.

Goldman: You spoke about meta-process control as a tool to help reduce the labor associated with dealing with assembly yield loss. Please tell me about the other subject the paper addresses—design for automation.

Borkes: A good example is the one we introduced in the paper. We looked at a product for opportunities to reduce labor cost by reducing labor content through automation. From a design point of view, components that require hand soldering, relatively speaking, are usually labor intensive. The components chosen in this case were standard right-angle pins (Figure 1).

In this application, they are usually found in groups of five per board and are soldered to small daughter circuit cards. The pins permit the small PCBs to be plugged-in to a motherboard, vertically, to save board real estate. The bare small circuit boards are fabricated in 200- up panels (i.e., using 1000 pins per panel. Normally what you do is singulate the panel and take five standard right-angle pins, put them in the board, and hand solder them. An assembly constraint is that when you’re done the five pins must parallel to each other.

We created a new pin. This pin was designed for total automation so you can print paste over the through-holes on the board, insert this pin, which has two legs on it, right into the paste; then, it can be sent through a reflow oven.

Now, you could ask, “Well, why don’t you just use the standard pin: print paste (paste-inhole), insert the pins and send it through the oven?” The problem is with a standard right-angle pin, the pins will rotate while going through the oven.

But these have to stay parallel to each other to permit them to be plugged into the sockets on the motherboard. This is a case study we use in the paper as an example of design for automation. You can redesign the boards to use the new, automatable pins as was done in the application; however, what’s better is if you design this kind of pin into the board to begin with. This allows the process to be automated without the cost and time involved in a board redesign. If you try and automate using a standard right-angle pin, you end up having to rework every board by hand soldering touch-up—it doesn’t make any sense. Even though you may have very good people doing the soldering, you occupy and pay them for the rework. This, of course, adds to the total labor cost of the product, making it harder to compete with low labor rate markets. That’s one example of the many we have done. But, you may ask, what about material cost of the new pin?

Here’s the punch line: This new pin costs the same, because it’s a stamping/coining process that we use to manufacture it. In other words, you are not paying any premium on the price of the material, which is important. One must always compare total costs, including material differences and non-recurring costs to see if there is a reasonable return—or, worse, any return—on the investment.

Read The Full Article Here

Editor's Note: This article originally appeared in the February 2016 issue of SMT Magazine.