Introduction

One of the most valuable courses I took during undergrad was called Art of Making (AoM). It was IDEO fieldguide-inspired course that heavily emphasized user-centered design. Design thinking is just like any other tool - there are times for it and times not for it. I’m not a huge fan of jargon about methods and affinity charts, but I definitely understand the importance of a deliberate design process. While it isn’t always the correct paradigm for a problem, it did teach me the importance of staying engaged client-side throughout all steps of the design process.

From sticky note ideation to ethnographic observation, having this incremental workflow that constantly prompted re-evaluation helped me build up creative confidence to picture something in my head and just start trying things. While not quite technically difficult, these projects help show the creative, non-STEM elements of engineering. I took the course freshmen year and can say with confidence that it really shaped my approach to engineering.

AnatoME - multisensory anatomical learning

This was the capstone project in AoM. We had been learning different prototyping techniques and design approaches, and then given full freedom to identify a problem and come up with a creative solution. Since my group was primarily pre-med and bioengineering, we tackled standard medical school learning pedagogy. Between team dynamic, progression timeline, and project enthusiasm, this was one of the most positive experiences of my time in undergrad!

Problem Statement

Anatomy is a cornerstone course that every health professions-aspiring student needs to master. However there thousands of parts to learn, which can be extraordinarily challenging. The challenge is compounded by the fact that the primary teaching method is with textbooks, an ill-suited medium for such a physically-rooted subject. Cheap models are tactile, but often lack textbook-level information density and context. While realistic models address the issue, but they are obviously reserved for more advanced study and can be prohibitively expensive. The goal of AnatoMe was to bridge this gap between 1-dimensional learning and literal cadavers in a way that would be more inviting and impressionable to beginners.

Prototyping

After brainstorming and some sticky note ideation, we decided that we wanted to make an interactive model the touched on several nontraditional learning senses. To start with a pre-prototype (pretotype), we took about an hour to make a quick proof of concept to simulate some functionality and get initial feedback.

This is a forearm model that mimicked the core functionality of what we imagined. I whittled some wooden dowels to make the radius and ulna, the team sewed some muscles together out of cotton and felt, and another member used littleBits to make a interactive light-up nerve. The muscles are removable and each part has a info card stapled on - the goal was to increase the tactility of models.

Ethnography

Armed with our prototype, we set up a table at the nursing school to have students try it out a give some observe, get opinions. We got all sorts of user testing - students who had taken anatomy, those who hadn’t, and even the odd professor over a couple days. Overall reception was positive, but by the end there some salient takeaways:

1. Electronics should be hidden or it distracted from the model
2. Muscle material should be more realistic if possible
3. We needed a fail-safe to ensure the user could put the model back accurately
4. Some users pointed out that beyond tactility, it’d be good to include something for auditory learners
5. Professors liked the idea but expressed doubt about how much information could be conveyed this way.

These were all important points that the team probably wouldn’t have addressed if we locked ourselves to the train tracks of the initial idea. And so we pivoted a bit to incorporate user feedback.

Adding Electronics	Silicone tactility

To tackle (1) we switched to a skull because it was A. cooler and B. had a convenient hollow brain cavity, perfect for hiding things. (2) I learned how to make custom injection-molded silicone muscles for a more rubbery feel and look. (3) I came up with the idea of magnetic connections that served 3 purposes - 1. easier removal and replacement than the previous velcro, 2. would force attachment in specific orientation, and 3. could be used to trigger electrical signals for (4 + 5) by adding a small display element that would read out the name of the removed part and provide extra information.

Final Design

The above essentially comprised the final design. Touch sensing and part removal was triggered with capacitive touch, and an arduino handled control and display. We went back to the nursing school to do some results testing versus an equivalent textbook learning procedure (and some more feedback of course). Although it wasn’t quite up to scientific testing standards, the results were significant enough to show that there is some merit to the idea of multisensory learning.

Final with program	Results

For the end of semester presentation, we put together this fun little video that storyboards everything you’ve read so far. None of us had any videography experience back then, but it definitely captured the spirit of what we wanted:

Bronchoscopy Trainer - lo-fi simulation tool

Bronchoscopy Trainer a client-based design project. The client was the WISER Institute, an organization dedicated to training health care professionals the practical simulation. This was one of the more iterative projects I have worked on - I came up with 3 lo-fi versions along with different sketches to talk with the client about. Each highlighted different key characteristics, and getting feedback on each helped give perspective into what exactly was wanted. Being able to communicate back and forth with the client so openly made this an incredibly valuable experience

Problem Statement

Medical simulation equipment can be prohibitively expensive. Although realistic simulators prepare experienced students well, sometimes simpler tools are needed to teach newer learners. Bronchoscopy simulators is such an area with a dearth of options for beginners.

Tom Dongilli, Director of Operations at WISER, filled in one of these options (above) to help introduce basic dexterity and control of bronchoscopes to newer learners. The goal of the project was to create another intermediate model that would help bridge the difficulty gap between the basic trainer and full body simulator.

Ethnography

I visited WISER several times to hear directly from Tom as well as observe how students typically learn with each of the existing options. I also used the bronchoscope myself to get first hand experience.

From the visits it became evident that while the project was fairly open-ended, there was an ideal trainer in mind. It should be:

1. extremely portable
2. have basic anatomical accuracy
3. incorporate easily varied levels of difficulty
   • in the dark or light environments
   • variable occlusions and deformations
4. have swappable end target panels
5. waterproof for fluid build up simulation
6. the solution should be economical

In comparison with the current beginner model, any addition of these would have been a fair improvement, so I had to prioritize which would be the most important features and hit them in that order.

Prototyping

mechanism ideas	form and feature ideas	potential model

I sketched out several pages of possible way to address the issues above. The biggest challenge was thinking of a solution that allowed for both instructor viewing and changeable obstacles in the airways. None of the pretotypes below answered both of these problems: the first didn’t allow for external viewing, the second and third were transparent, but as a result I couldn’t swap out occlusions.

Option 1	Option 2	Option 3

The answer actually came relatively simply - vacuum molding could provide a way to shape clear plastics like PETG into a lung shape. Moreover, it was actually preferable to have the airway bisected (which is required by vacuum forming), since this would allow for easy internal access - then a clay or poster-tack could be used to place custom, easily cleaned, occlusions.

The model above is a quick CAD mock-up of the final concept. The red object is the airway on supports - made by printing out bisected halves of the airway and vacuum forming over them. The box around could be simply laser cut. This solution addresses all of the initial goals, better than I had hoped.

Final Design

Below is the latest prototype, created using the steps outlined above. You’ll notice that instead of plastic supports, I have a 3d printed base support due to the planned dimensions being too thin. As a result, tabs with magnets were used for registration, which I hope to phase out.

Initial array	Final Optimized

The client was very satisfied with the concept, and we have begun moving forward to make a more polished design with quality materials. Although Co-vid 19 has slowed progress, I’ve been able to use the time to update the model and find an injection mold company that is willing to produce parts for low volume production. Updates coming soon.