Many of you were curious about the “snap together construction” and “one minute assembly time" of this years gearboxes. In preparation for our senior project presentation and our final report, we conducted a time study and efficiency analysis of the gearbox assembly in line with some of our industrial and manufacturing engineering requirements for the project. The setup shown in the video was created to mimic a production environment, where parts and hardware are stored in replenishable bins, or on angled racks, feeding a new part in a specific orientation both aiding and speeding up the assembly of each finished gearbox. This is in no way a representation of how these were assembled in real life during the six-week build. This is to show how and why some of the design decisions were made in an effort to decrease assembly and dissasembly/maintenance times.
Here is the video: Gearbox Assembly
Our initial assembly times, prior to any right hand left hand part placement optimization gave us assembly times in the two to three minute range. After dozens of tests, optimizing the part placement and the assembly procedure, we eliminated many of the awkward hand swapping and extended part reach slowdowns. Our first few tests with an optimized parts placement layout gave us times in the high one minute range. After a long and slow learning curve, our first assembler was able to finish in one minute 20 seconds.
After further refinement of the layout and the assembly sequence, we were able to meet our estimated one minute assembly figure. Unfortunately, this only happened after nearly 50 previous efforts. While a continual decrease in time yielded a rather typical learning curve, the final one minute time is below what we have concluded to be ‘optimal’, being somewhere in the neighborhood of 130% efficient for a typical worker on an eight-hour shift.
This efficiency simply means that in normal production scenarios, the worker would not be assembling these one per minute. An average assembly time near 1:20 is a more reasonable expectation after our limited testing.
Many of you might say “he cheated” or “that’s not very fast” but please remember, this study was done more as an example for what would be done in a real world production scenario. While it IS possible to be done faster, the ability for most people to maintain such a speed for a typical 8 hour workday is near impossible. Our original estimates of 1:00 had already taken into account some of the tasks that would be done prior to final assembly. Ex. bearings being pressed into plates. The setup could have been different, but we chose to make the assembly as comprehensive as we could, while eliminating the mundane and slow tasks such as bearing pressing.
Some of you may notice the lack of hardware. This is by design. The removal of all nuts from the gearboxes saved nearly two minutes of assembly, even while accounting for the additional time required to press in threaded inserts. The removal of keystock and keyways eliminated an additional one minute from assembly as well. While these things may increase the initial manufacturing time and costs, the savings in weight, increase in strength, and decreased assembly/disassembly/maintenance times was determined to be well worth it. The reduction of the total quantity of parts as well as the design of critical components such as the output shaft also allowed the assembly time to decrease and enable quicker and easier removal/tracking of parts for disassembly and maintenance. Spacers were removed or integrated into existing components when the weight savings difference was negligible.
Overall, this gearbox was designed to be as light as possible, with as few parts as required while being simple and quick to assemble and disassemble using only a handful of tools. We feel we have accomplished it well, but as always, there are improvements to be made.
While this is only the first of a few posts we plan on making with regards to the gearboxes, we would like to thank all of those who inspired us as high school students to pursue engineering who also sparked many of these ideas and created many solutions that we borrowed throughout our time in FIRST. Andy Baker, Paul Copioli, Joe Johnson, and Glenn Thoroughman to name a few… there are many more out there… but we wanted to recognize these guys specifically. It is important to create your own ideas and to continually improve existing ones, but more important than that, you must remember and acknowledge those whose footprints you follow in. These guys really helped us out over the years, both directly and indirectly… we thank them and all the others we haven’t mentioned! We hope to one day be as good of engineers and role models as these folks are.
Please feel free to ask any questions, too!