So, this happened: Marketing sold a nonexistent product based on a 3D outer surface model and a glossy product features brochure. Orders were taken. The market introduction date was set, plus or minus a few weeks.
The product was a slight improvement over the competition, with a significantly lower price tag. All its magic was in the software. As new orders poured in, engineering management had doubts about meeting both the introduction date and the required volumes. They hired me. I am given a week to get everything organized and then the clock starts.
The Big Boss holds an all-hands meeting for this project, introduces me, gives the usual "rah-rah" speech and leaves. The head of engineering heaps me with as-yet undeserved praise and says that I'm the engineering manager on this project because he is going offshore to interview new production sources and to transfer some existing products. He gives me carte blanche management authority for the term of the project.
He gives the teamwork speech, ending with the semi-hilarious, "there is no I in team" delivered with a straight face. This is a fast-track project. I am either going to deliver it within the timeframe or am expected to fall on my sword with a much smaller invoice for my effort. I am working with a bunch of people who are used to pointing at someone else when something is late or goes wrong. I tell the team we are all going to work together, but differently. I repeat that there is no I in team but, more importantly, there is no u, either. There is no end to this project for anyone on the team until we are in full production.
I give them my mission statement: "All blessings flow from the shipping dock. If you're not working directly to put quality products in volume, on time on the shipping dock, or assisting someone who does, you shouldn't work here." For this project, we will hang together, or we will be hung separately. Fast tracking is not hard. However, it does not tolerate slackers.
I start with the designers. Since they already knew what the outside looked like, I ask them about the interior: How will they attach cooling fans for the electronics? Will the PC boards be flat or stand on end? How will maintenance be done? Are we making a five-sided box where we can pull the entire front or back out for maintenance/assembly, or are we making a top and bottom that will fit together? Since this is a snazzy upgrade to another product, I tell them to look at previous designs and see what they can steal and copy into the new designs.
The first bomb drops: "The previous case part designs haven't been updated," I'm informed. "We write the change orders, and then tooling/purchasing/quality implements them. They go through an inspection process, but nobody updates the CAD drawings. These are all ‘red-line’ changes."
"Who's responsible for doing the updates?" They mention Bob, who is not in the meeting because he is out for an extended lunch talking to a sales rep. He is waiting for his retirement (slacker). I tell the designers to get me a list ASAP of the parts/assemblies that do not meet the print in order of their importance to this project. I will get them updated prints in less than a week. This comment is met with the expected eyeball rolling and muffled, "Good luck with that (sucker)."
I move on. I told the designers to do as much huddling as possible with the electronics and robotics engineers. When their designs are semi-ready, they are to print them out and hang them on the conference room walls. I told the tooling, quality, electronics and robotics people to mark them up with anything that wasn't doable once they were on the wall and to remember:
- There is no such thing as zero draft.
- A CpK of 1.6 cuts the specified tolerance in half. If you absolutely need that precision, specify and justify it (along with two alternative designs that accomplish the function but don't require this precision).
- There is no such thing as a high polish unless the polisher was stuffing recreational pharmaceuticals up his or her nose. There are polishing standards. Find them. Outsides are cosmetic; insides are not. Talk to the texturing house and find out what they need.
The designers must meet their completion dates; this meant that every engineering function had to sign the prints as being acceptable so that tooling could build molds, jigs and fixtures based solely on the designs. Critiques were to be done the day the designs were hung up.
In some cases the information was not available—i.e., they couldn't specify the dimensions of bosses and screw holes because they didn't know what brand of standard component purchasing they would buy. "Leave it blank," I told them. Highlight this area and the tooling folks will leave it steel safe. When we get the info, we will add it to the design and cut steel.
"What about engineering changes?" The designers get only two changes. First, when the tooling is 50% complete they process a massive ECO incorporating all the changes to date. That will minimize the cost and timing delay. Second, the last set of changes will be just before final fit-and-assembly. This also allowed the tooling engineers an advance look at the upcoming changes so they can plan for it and not delay the schedule.
We will not wait for parts to build assembly tooling. If you have faith in CNC, you do not need production parts. Gin up a few prototypes and use them if you have to.
"But what if marketing asks for enhancements?" I tell them, "If we can fit it into the schedule, we will. However, if it delays the schedule in any way, they will explain it to the Big Bosses and explain the late deliveries to their customers. With upper management's approval of a later delivery date, no problem. I've arranged for marketing's bonuses to be a function of on-time completion of the project and the accuracy of their forecasts. Over-tooling wastes money that could fund other projects. Under-tooling means we cannot keep up with demand and we lose sales. As you would expect, I'm no longer on marketing's Christmas card list, but they will get over it.”
Now, the tooling guys and gals. Since we already knew the outside profile, the designers immediately decided it would be a top and bottom assembly and the bottom could/might need slides on three sides. We had all the info necessary to begin the tooling program.
I told the tooling engineers to partner up with purchasing and find no more than three tooling/equipment manufacturers within 100 miles of their location with whom they had a successful history. The closer the better. Since my client was near Chicago, it would not be a problem.
Their first assignment was to figure out, based on market estimates, what kind and how many tools were required. They would then go to the suppliers and explain a “no-fault PO”:
Phase 1: We will pay for all long lead items before the designs are released (mold bases, materials for construction of the assembly machine's base and so forth).
Phase 2: When preliminary designs are released, we will pay for the design of the tooling and equipment.
Phase 3: Based on the tool and equipment designs, we will pay for the initial construction of the tools but not the final machining required by a finished part design.
Phase 4: With final designs, they would be released to do the final machining/polishing/construction.
Phase 5: Tool/equipment qualification.
This project was split into phases to get the suppliers paid as the project progressed and not hold back monies they were owed “just to keep them honest” (which is the dumbest way to treat a supplier whose work will get you your paycheck). Purchase orders were payable n-10 on completion of each phase; because this was a fast track, I wasn't going to let invoices age like good whiskey. Phases one and two could be done in parallel and didn't need to be done sequentially.
This was a "best-possible-one-time-price-and-timing" bid. I told the buyer/engineer teams there would be no hanky panky in vendor selection: You work with the vendor who delivers on time, not the guy who offered you season tickets to the sport of your choice.
I gave the vendor an electronic spreadsheet progress report I'd written. The vendor would e-mail me one column per week that I would splice into an overall progress report and e-mail to him with my comments. Along with the spreadsheet data, he would send pictures and/or video showing proof of his progress. We would meet electronically over the internet. If things were on time, the meeting would take five minutes. If things needed a discussion, all concerned parties would be present, prepared and on topic with written reports, audio visuals, etc. e-mailed out ahead of time. That meeting would last no more than an hour. Meetings are to share information and make decisions. They need to be on topic and as brief as possible. If you want a sermon, go to church. This avoided taking all-day vendor visits, long (expensive) lunches and other time-wasting silliness. If the vendor could not meet these requirements, we wouldn't work with him.
The morning was spent finding potential vendors. It took an entire afternoon to train the buyer/engineer team how to use the progress reports. The next day I put them on the road for a face-to-face visit with their future suppliers carrying the marketing show-and-tell 3D models, parts from similar products and the most recent sketches and drawings available. I expected hard copy quotes in five business days after they received the five RFQs. I explained some of the final machining/fabrication (phases three and four) might be difficult to hard quote. We'd do that on a time and materials basis initially. When they got back from their road trips, I expected fully written RFQs and as much information they could give to their “chosen” supplier e-mailed within three business days. When they baulked, I reminded them all RFQs and POs are usually fill-in-the-blank word processing forms. They were to place phases one and two immediately and phase three as soon as possible.
I tracked down Bob, the ECO guy who didn't like paperwork. His explanation was he was too busy to process ECOs. I told him I had just made his life simple. No more vendor lunches or anything else that took time away from his desk or this project. I handed him the list the designers gave me of the most critical parts they needed so they could copy portions of the design into our new product.
"You are to finalize (read: update the CAD designs so that they reflect the actual part) at the rate of 10 ECOs/day minimum, beginning at the top of the list. Start with QC inspection reports. I want the entire list done in five days. E-mail the final designs to the designers and myself. As soon as you finish with one part, go on to the next. If you can't (won't) do it, the guy or gal who replaces you will. I'll get you transferred to the ISO compliance section where they'll bury you in paperwork." Problem solved. I didn't care if he ended up working 16 hour days. Yes, I have an issue with slackers. I made a mental note to rat him out to upper management so that all the other non-processed ECOs would be done after the project was complete.
Everyone was off to the races. I went back to the designers. This product had a small keyboard and LCD display. They had always used slides to create the holes. I asked them if there was any reason we had to have both of these square to the viewer other than because “we’ve always done it that way?" No one could come up with a reason. I suggested for as many sides as feasible to design the part for pass-through cores. No moving parts. With enough draft, the core and cavity sides could kiss each other and create the openings. It was a light-bulb moment for them. Since we knew there were going to be several vertical PC boards that would require slots or something to hold them in position, I then asked the "Sparky" engineers to design a motherboard we could screw into the base and affix the vertical boards to that.
The objection came back that the bosses that accepted the assembly screws never lined up with the PC board. I told them to tell the PC board geeks to leave room where the screws were supposed to be. Since PC boards were all cut individually from a larger sheet of epoxy/fiberglass, we'd tell them exactly where the holes in the bosses were after the qualification run and it was a simple task to modify the CNC router/waterjet system as opposed to moving a molded hole a few thousandths of an inch. The designers scurried over to the Sparky lab.
The clock started. Fortunately, the vendors were picked, the POs were written with phases one and two released and three to five awaiting release. Magically, ECOs were processed, and designs updated. The designers did the cut-and-paste trick, plucking the geometry for switch connectors, power cord attachments and so forth from old products and marrying them to the new designs. Then they staged a “raid” on purchasing to get the design specs on all purchased components that would attach to the molded parts. Being forewarned (by me), the buyers announced 90% of the purchased components was stuff they already had in stock for other products. What was left was a few parts from catalogs already dimensioned; through a “website grab,” they could be downloaded in any of several CAD formats. (My talk about standardization was paying off.)
I told the troops I was going back to Denver, but would be in touch daily until things were running smoothly.