Last week marked our biggest research effort to date, with Lisa, Maggie and Kirsten travelling to the Bay Area to visit Specialized Bikes and Ames, and Adam and Derin venturing out to White Sands. Here’s a breakdown of our observations:
-
Specialized Bikes: 2 test labs, 1 machine shop
-
Ames: 2 Arc Jet test labs
-
White Sands: 3 machine shop, 2 test labs
We also interviewed a Stanford anesthesiologist and had several impromptu interviews at Ames and White Sands. Our research reinforced earlier findings about individuals’ use of memory aids during procedure execution, the variability in individuals’ tool tracking techniques, and the use of “assembly line” processes. We also observed the challenges of collaborating over large distances, saw the inefficiencies created by dependent workflows, and discovered instances in which incorporating technology into one’s work is challenging or impractical. When we return from spring break on 3/17, we’ll continue consolidating our findings (so far we’ve consolidated the lab scientist domain).
Our trip began with observations at Specialized Bikes. We watched two technicians perform different tests – one to ensure that the handlebars on a road bike could bear load at different angles, and the other to verify that a road bike’s frame was stiff enough. We also saw a machinist cutting a metal bike part using a lathe. One interesting finding was the techniques a technician used to track and organize his tools. In addition to placing his tools inside labeled or unlabeled drawers near his workstation (like many individuals in other domains did), he marked them with a thick blue stripe to signify that they belonged to him. “That way, others will know to return these tools to my workstation.” We also saw the same technician use memory aids such as marking handlebars with the letter, E, to remind himself which test to perform. Finally, we observed a technician run the same test on multiple frames before moving on, instead of performing all the tests on one frame. It would be interesting to know how individuals who employ “assembly line” techniques on the job group which task(s) to perform in each round.
(Specialized Bikes in Morgan Hill, CA)
We also observed two Arc Jet mechanics at Ames pumping water through tubes to test for leaks, to prepare for running a future test. One interesting observation was the challenge of collaborating over large distances, which is a problem that astronauts face on the ISS. The Arc Jet mechanics worked together but had trouble hearing each other since they were each located at different parts of the lab. Thus, they coordinated their actions by yelling across the tubes, which increases the likelihood of mishearing one another and making mistakes. Moreover, we saw the inefficiencies created by dependent workflows. For example, the Arc Jet operator had to leave the control room and individually ask each mechanic if everything was ok before proceeding. Finally, we witnessed NASA’s “safety first” culture firsthand, which was evidenced by the operator’s claims (personal and facility safety are her top priorities, and she always has a backup plan) and what we saw: emergency stop buttons all over the control room, a control screen in the control room that’s mirrored into the Arc Jet room–giving workers more control and information, and the practice of operating in only one control room per day to avoid mixing them up.
At White Sands, we observed 1 machine shop and 2 test labs and saw similar issues encountered at Ames. Again, we witnessed the inefficiencies created by dependent workflows. For example, a lab technician at the hypervelocity chamber relied on a coworker to verify that he was done using a box so he could put the lid on but could not reach his coworker through multiple attempts (calling on phone, paging, calling on intercom). Eventually the coworker walked by, oblivious that the lab technician had been trying to reach him. We also observed the challenges of collaborating over large distances. For instance, a machinist who was making a part (sabow–sp?) for the hypervelocity chamber had to drive to the hypervelocity chamber to verify that the component fit.
A few of our White Sands observations did not manifest at Ames, however. For example, one of the White Sands machinists worked with tiny materials that he stored in bottles because of their size. Moreover, we saw different techniques for tagging materials. For instance, at the machine shop/calibration lab, almost everything was stamped or tagged, while nothing was tagged at the hypervelocity chamber. These observations showed us how techniques for storing and tagging materials can vary based on their size and type, and likely other factors such as weight, value, etc.
Finally, we interviewed an anesthesiologist about his tool tracking techniques and procedure execution process. One interesting finding was the difficulty of using technology to verify that he had given a patient the correct drug(s) and dosage(s). The anesthesiologists current methods include manually verifying which drug(s) he administers and monitoring a patient’s vitals on a computer screen. Technological verification methods, such as a system that can detect whether the correct drug and dosage was administered, don’t exist because they’re not scalable. Patients require different drug(s) and dosage(s) based on individual traits such as body weight, age, drug use, etc, and accounting for this myriad of possibilities is nearly impossible. Another key finding revealed that using technology to solve a problem is not always appropriate. For example, anesthesiologists currently reference a crisis management book, the Emergency Manual, during procedures if something goes awry. While delivering the information in traditional book format may seem outdated, technological solutions such as audio dictation don’t work in a noisy environment, nor do tablets/other touchscreen devices if the anesthesiologist is wearing gloves or carrying tools. This reinforces the need to consider the individual’s work context when designing solutions to a problem.
(Emergency Manual, created at Stanford)
Stay tuned for more info when we return from spring break!
Finishing up the lab scientist flow models consolidation
Helios 