A lot has happened since I last wrote a retrospective about Nolop, Tufts’ makerspace for the entire university community. Let’s look back at the last 2 years in Nolop.
Before the pandemic
In 2019, Nolop grew increasingly popular. For the last few weeks of the fall 2019 semester, Nolop was packed every afternoon, routinely drawing crowds of 50+ people from 4 PM until 8 or 9 PM. The laser cutter was running continuously most of the day and late into the night. Our four 3D printers also ran close to 24 hours per day. With grad student Elliot Pavlovich supervising the red zone several evenings per week, use of the table saw, drill press, and band saw increased as well. Another student, Mandy Rosengren, helped us set up a wearable tech area with a sewing machine and electronics for embedding in garments and accessories. We added a jump shear and brake for working sheet metal and a second soldering station to support more users.
In the winter of 2019-2020, we bought 4 more Prusa 3D printers to expand our capacity. I built a set of shelves to hold all the printers efficiently, and Eric Wu devised a mounting system to hold the Raspberry Pi print servers. We also added a Formlabs Form 2 printer for high precision prints.
Over the course of the year, the Nolop community grew tremendously. We had more users every week, and we developed a core of around 30 students, who, in addition to my 20 student workers, knew how to use most of the equipment and helped orient new people to the space. This community is key to why Nolop has worked well— instead of the staff being a bottleneck of expertise, students learn heavily from each other. This kind of peer learning and exploration also prepares them for professional settings, especially in innovation work, where nobody knows the answers yet.
More classes using the space
During Nolop’s first full academic year of operation, students used the space for class projects more and more. All 41 mechanical engineers in the senior class built a wooden chair as a warm-up project for the Senior Capstone course.
We also saw increased curricular usage from biomedical engineering, various robotics classes, and some sections of ES 2, Introduction to Computation in Engineering. Civil engineering students built wooden bridges for their structural engineering class.
Clubs and individuals using the space
In addition to class projects, we also saw increased usage by student clubs and individuals. The Tufts Robotics Club held meetings every Friday in Nolop. The space/rocketry club, SEDS, built rockets in Nolop; a weather balloon they launched made it within 20 miles of the Canadian border. Tufts Make also used the makerspace for meetings and project work.
We continued to hold recurring events, like the Nolop Haunted House (4th annual event already coming up this fall!).
The pandemic arrives
On March 15th, 2020, Nolop closed due to the pandemic.
For the end of the spring semester and through the summer, Nolop was truly closed. We started a Slack workspace so everybody could keep in touch. In the early days of the pandemic, we also 3D printed filter adapters for health-care workers dealing with Covid patients.
Throughout the spring and summer, I livestreamed on Twitch from the red zone as I built a crosscut sled and a heavy-duty stand for our sheet metal brake. I also streamed electronics teardowns, the most popular of which focused on an old projector, a delightful mix of electrical, mechanical, and optical engineering constraints.
Remote fabrication
As we moved into the academic 2020-21 year, most students were back in Medford/Somerville, but watching classes on Zoom. Almost all of the Nolop staff returned to campus. We switched to a new model, in which remote students could submit jobs for fabrication through our Slack channels. The Nolop student workers worked, masked, in the makerspace in shifts. As fabrication requests came in, they would load the designs into laser cutter, or set up the 3D printers with the right filament and make sure the beds were clear for printing. After the parts were made, they would leave them on a table in the hallway outside Nolop for pickup. This is how we made it through the Covid winter.
One unexpected benefit to having the makerspace run just by a skeleton crew during the pandemic was the strong community sense developed among the 14 student workers. They were in the trenches together; I expect that many of them will be friends for life. I’m enormously proud of what they accomplished and grateful for all their efforts.
In addition to lasercutting and 3D printing, we also provided soldering services, as well as help for general stuff like, “Can you drill a hole in this board for me?”
Despite having to send everything through Slack, people still built remarkable stuff, like Zachary Goldstein’s air engine, below.
Estimating remote usage during the pandemic from logs
While Nolop was operating under the remote model, we have a couple of ways of estimating usage. The first is that for all the jobs that were submitted by people outside of Nolop, we have Slack conversations. Here’s what that data looks like for the last year.
We have another source of data, which is the laser cutter PC. Every time someone sends a job to the laser cutter, the PC logs the job in an event log. Here’s that data, overlaid onto the Slack data for the laser cutter.
The peaks of the Slack data and the PC data roughly line up, but the PC data is, on average, around 2.75x higher. If we use this ratio between Slack and PC data as a correction factor for 3D printing and soldering, we can get an upper bound on how many jobs the Nolop staff completed over the course of the two semesters: 8352, or an average of around 35 per day. My suspicion that the true number is more in the range of 5000-6000.
| tool | Slack | PC event logs | scaled estimate |
|---|---|---|---|
| 3D printers | 1830 | N/A | 5036 |
| laser cutter | 1093 | 3008 | 3008 |
| soldering | 112 | N/A | 308 |
| total | 8352 |
Class use during the pandemic
The total number of courses that used Nolop, tabulated below, decreased during the pandemic, but the total number of curricular projects probably increased. Before the pandemic, some of the usage we saw was on open-ended projects, where professors gave students the option to build something if they wanted to; many students chose not to. Over the last year, we saw several classes use Nolop as a required part of class work. For example, every sophomore mechanical engineer was enrolled in ME 21, Dr. Wendell’s Mechanics class; all 85 of them lasercut some kind of mechanism through the Nolop Slack channel. Similarly, most of the ~60 junior mechanical engineers soldered their circuit board projects through the Nolop Slack.
| Course | Instructor | |
|---|---|---|
| BME 66 | Engineering Design Process | Janet Krevolin |
| EN 1 | Simple Robotics | Ethan Danahy |
| EN 1 | Inventing Smart Toys for Kids | Briana Bouchard |
| ME 21 | Mechanics II | Kristen Wendell |
| ME 30 | Electronics | Brandon Stafford and Kristen Wendell |
| ME 31 | Controls | Rob White |
| ME 70 | Instruments and Experiments | Felix Huang |
| ME 74 | Senior Design Project | Jeff Guasto and Felix Huang |
| ME 116 | Composites | Michael Zimmerman |
| ME 134 | Advanced Robotics | Matthew Woodward |
| ME 140 | Inventive Design | Gary Leisk |
| ENP/ME 193 | Assistive Design | Dan Hannon and Gary Leisk |
| ME 193 | Microcontroller Programming Projects | Brandon Stafford |
| MUS 63 | Electronics Musical Instrument Design | Paul Lehrman |
Non-curricular usage during the pandemic
Non-curricular usage dropped dramatically during the pandemic. The Nolop staff still worked on personal projects, but the vast majority of what we saw come through the Slack channels was for coursework. (The Robotics Club did continue to build robots remotely, but we can’t take credit for that.)
There are more images of the ceramic printer available on the web.
We also saw various research groups and other entities at Tufts using the space. In support of Tufts covid testing response, Tufts Technology Services, led by Tufts CIO Chris Sedore, recruited some engineering students to build devices to make virus testing work more smoothly.
The first machine adapts a label printer to peel labels, with the eventual goal of applying them to the sample tubes without everybody smearing their biological juices on everything. The second machine used a suite of cameras to image every sample tube, in an effort to use machine learning to identify the cause of a small number of inconclusive tests.
Materials usage
We were previously averaging around 40 kg of printer filament per semester. That average held through the pandemic, more or less. We bought 142 kg of filament over the last two years, but approximately 10 kg were discarded because a certain brand tended to clog our extruders. If we omit the half of the spring 2020 semester when we were closed, that’s about 38 kg per semester, roughly on pace with our previous usage.
The Nolop Store also continued to resell materials at bulk prices at a brisk rate, around $1000 per semester. These are materials, like acrylic for lasercutting, or batteries for electronics projects, which are cheap in bulk, but expensive in small quantity.
Workshops
Before the pandemic, we ran workshops to teach people different fabrication techniques and draw people into the space. We couldn’t continue that with the pandemic, but we did partner with Tufts Design For Social Good and Cecropia Strong for a virtual weekend hack-a-thon called TAThack. I met with students over Zoom to give them design and fabrication advice as they designed assistive devices for a 9-year-old girl with a congenital spinal cord injury. I found it particularly rewarding to help non-engineering students who were totally new to the field, but bravely pushing forward.
Safety record: excellent
According to our safety log, before the pandemic, we were averaging 7-8 band-aids per year, with no serious injuries so far. Since the last report, we have had only one potentially serious safety incident, which was a small fire in the laser cutter. The laser typically produces a small flame when cutting wood. In this case, the laser was set to cut thicker material than was actually being used. Additionally, we suspect that some bits of material had fallen into the grid of the laser bed, where it was susceptible to ignition.
The person operating the laser cutter noticed the fire immediately (good job, Teddy), and it was extinguished quickly by another staff member (thanks, Jeff). No substantial damage was caused to either people or the laser itself.
Summer 2021: equipment maintenance
The Covid testing program at Tufts is suspended for the summer, and most of my student workers graduated anyway, so Nolop is closed to the community for the summer. This has left me with the opportunity to fix a lot of stuff that broke over the last year.
The honeycomb bed of the laser cutter has been replaced. The upper left corner, near the origin of the laser’s coordinate system, had been heavily ablated. In the future, we should remember to flip over the honeycomb before it gets destroyed. The cost of this repair was also low, around $150 including shipping from Arizona.
I just spent a week repairing our 3D printer farm. Three of our eight printers needed their extruders replaced– one was totally destroyed; the other two were limping along, but with reduced print quality. I found a wide range of inventive hacks used to keep the machines functional; I tried to restore them all to factory condition. We’ve now had the printers for an average of 2 years, and the total spent on replacement parts across the eight of them is in the range of $500-600, slightly less than the cost of one new printer.
The future: waterjet and beyond
Thanks to the diligent efforts of Jeremy Kanovsky and Tufts Robotics, we have a new Wazer waterjet cutter that has just arrived; I hope to get that set up and running this fall. It should provide a beautiful workflow, in which designs are rapidly prototyped out of wood or acrylic on the laser, and then the validated designs are cut out of metal on the waterjet.
I’m also working on a new open source robotics project: an autonomous vacuum bot for the red zone– like a Roomba, but big like a Shopvac. We had a first version of this built by an ME capstone team in 2019, which validated the concept. My hope is that I can get a basic machine working this summer and pass it off to students for improvement. (The closest I’ve seen to this commercially is the Makita DRC200Z, which I can’t program myself at all.)
I’m a little nervous about how the fall semester will go with relatively large groups of students working (masked) in the makerspace together, but Tufts’ Covid response has been outstanding so far. I’m immensely proud of how we reacted as a community to tough, changing situations. I think we’ll do well.
On to the post-pandemic era, at last, we hope!
Brandon Stafford
Manager, Nolop Makerspace