xcv Flashcards
Customer Obsession
-Automatic tool change of probe tips. Machine has two FANUC robots on it to swap tools. 8 different probe tips live on a rack 12 feet up in the air on a machine. They need to be swapped daily for cleaning. I could have just made the customer climb up there on a stepladder every day to change the tips, but instead I designed a custom robotic interface for this rack so it could be lowered down to ground level and handed off to an operator.
Ownership
-7 am email of X axis crash in Renton, my (relatively unimportant) part is mangled. Went down and worked through the day and past midnight pulling off hundreds of feet of 50mm rail, stoning damaged surfaces, moving damaged rail to non-production side of the cell and good rail into production side, reentered production just after midnight
Invent and Simplify
-New technology “servocreel” requires new spool retention mechanism. Needs to throw 130 lbs of force to hold the spool with friction, but must be quickly and easily locked and released by hand. First design was done by a senior engineer from outside the group, very poor ergonomics, need to twist large knob four times to lock the spool, high torque hurts wrist. My design uses a pull/push action. Long stroke of pulling action means much higher force is delivered from minimal operator effort
Leaders are Right, a lot
Worked with a controls programmer to develop an automatic hole probe. Characterizing automatic inspection tool against their proving rings to demonstrate accuracy of the system. Accuracy requirements within +/- .0002”. Hole diameters reported to the HMI within one tenth of a thou. I’m collecting thousands of data points, something starts to bother me; data doesn’t seem to follow a normal distribution. Low and high results are suspiciously common. I suspected bad rounding in the code. Programmer told me nothing was wrong, took it up the chain of command, and found a very serious error.
Learn and be curious
-Learned ANSYS in my spare time, end up leveraging that to cut 30% of the weight from AFP head structure, through analysis and optimization.
Hire and Develop the Best
-New pneumatic engineer has been brought into the AFP group. AFP is the most prestigious group at EI, challenging work and demanding managers, highly complex systems. To contrast, pneumatics is the least desirable specialization. Working with this engineer, I quickly realized he was highly smart, ambitious, and hardworking. He was enthused to working on this important R&D effort and really threw himself into it. Realizing his talent and commitment, I challenged him to radically improve the pneumatic air distribution on our heads. We worked together, incorporating pneumatic manifolds into the structure of the head, passing many air passages through solid, aluminum structural members. This eliminated over 50% of the tubing our heads previously had. Lesson I learned here is, when you get lucky with good talent, give them the biggest challenge possible.
Insist on the Highest standards
-Working with a junior engineer on new dancer design, adapting the 8 tow dancer to the new, 16 tow “servocreel” technology. He wants to use the traditional 8 tow springs, which will be considerably easier, allowing reuse of certain components. However, a different spring exists that has tiny (5% ) better tension characteristics for servocreel. I advised him to use the optimal spring despite the extra effort and small benefit, because this design was likely to be propagated and when it comes to our process, no amount of performance is too small to reach for.
Think Big
- In-process inspection is an example. AFP is a very expensive, time consuming process. Typical inspection may involve 6 QA techs walking the part, manually inspecting every tow. To make it viable for the next Boeing airplane (smaller, high volume) it’s important to eliminate manual inspection. In support of this I altered the tow tensioner to a magnetic encoder from a simple analog position sensor. This increased accuracy helped enable “real time in process inspection” by allowing software to precisely locate tow ends. That sort of change could make our process viable on high-production rate aircraft in the future, by eliminating the need for manual inspection steps.
Bias for Action
-With an important customer demo only days away, a new design of mine undergoes a pretty spectacular failure. Spools fell off the machine which is dangerous and makes a big mess. PM asks me if he needs to delay the demo to give me time to design a fix since a repeat of that in front of the customer would be unacceptable. I work into the night testing out three redundant improvements to the design. By the next morning I tell the PM to go ahead with the demo, my machine won’t drop spools again. Successful demo results in a sale, I haven’t dropped a spool since.
Earn Trust
-As tech group lead, I’m often working with less experienced engineers as they design their first assemblies in the group. One technique I always use to build trust is to take blame for their mistakes, but give credit for any successful ideas. This builds trust both up and down the chain of command. Newer guys trust you will give them credit and won’t throw them under the bus, leadership sees that you take responsibility for mistakes made under your purview. This reinforces the perception that design under my umbrella is held to my high standards, even when that work has been delegated.
How did you handle a Difficult Coworker
-Introduction of my material spindle “servocreel” was a highly political development. A sr. engineer hand picked by the owner had been brought into the group to invent this device, and had failed utterly. We had a design conclave where many managers and high ranking engineers and people brainstormed thoroughly but came up with no good options. I later came up with my concept and I was totally convinced it was the best I had. To maintain trust with my Josh and my group, including the aforementioned sr. engineer who’s design I wanted to replace, I presented my idea first to him, asked for his advice, incorporated details based on his feedback before moving then to the rest of my group and finally department leadership.
Dive Deep
-I’ve built over 20 AFP heads over the past three years, more than just about anybody. Although much of the assembly is delegated/outsourced, on every single head I personally torque, retorque, and antisabotage paint safety critical bolts. Our department uses zinc plated fasteners, which can rarely undergo hydrogen embrittlement. Torque the bolt, seems fine, then down the road the fastener head pops off. By waiting and rechecking 24 hours later, you eliminate this possibility. I use a particular color paint pen which serves as a visual indication that a given fastener is as safe as possible.
Have a backbone, disagree and commit
-SAL ultrasonic tank full of solvent, was intended to live on the machine per the customer spec. Probe tips would be robotically plucked from the machine and installed into the probe. This sticks out as the first time I was asked to implement a really, really bad idea. My immediate manager wanted me to flesh out a design. I did, but also presented the massive downsides of this idea to the group, the idea was shelved.
Deliver Results
-The challenge of building an 8 tow servo head. Small, niche product, only one in our lineup still using last-gen spindles. Due to inspection capabilities of servocreel necessary to invent a solution. Huge multidisciplinary design/packaging challenge required total redesign of the servo spindle with new gearbox/motor to solve. Additional stacked IO rack, dencest electrical cabinet in department history by far. This allowed us to roll out the huge performance/inspection/reliability advantages of servocreel out over our entire lineup.
Why Amazon?
-Comes back to amazon’s principles. I feel amazon is far more grounded, dealing with and fixing real world problems that have had huge positive impacts on me. Story about the mall/buying jeans. It’s my own experience as a customer over the last 10 years that has convinced me that amazon is well positioned to grow, due to it’s focus on real, everyday problems that consumers face.
