Mechanical Engineering Capstone:
TerraRover
Fall 2022 - Spring 2023
I served as Team Lead for this project, directing and delegating work as prescribed by our correspondants at the AREN Project and GLOBE (NASA). This project was the culmination of all things learned over my four years at the Boston Univerisity College of Engineering.
Special thanks to Andy Henry (AREN, RESA), Geoff Bland (GLOBE, NASA), Dr. Anthony Linn, and Dr. Caleb Farny.
Fig. 1. Solidworks model of the TerraRover with UHIP 2.0.
Fig. 2. This is a snippet from our "how-to-build" tutorial video showing the TerraRover driving using RC.
Abstract:
TerraRover is the name of the device our team of four built in conjunction with the AREN Project and GLOBE, the latter of which is funded by NASA's Science Activation Program. The purpose of this remote-contolled vehicle is to be able to collect data pertaining to the Urban Heat Island Effect (temperature, wind, CO2 levels, etc.). This device was also meant to be a teaching tool for high school-aged students, the idea being that students would be able to build this Rover, and collect Urban Heat Island data for NASA. Our job was to create a concise manual for the build and operation of this rover and improve designs as we saw fit, targeted at high school aged students.
​
As a student and budding professional, this project was extremely rewarding to work on because it tested my hard skills, specifically with design and prototyping, and especially tested my soft skills in effective communication and teamwork. I couldn't be more excited to showcase the work we put into this project -- I am proud of it, and the TerraRover team at AREN and GLOBE are also proud of the progress we made.
​
For the sake of brevity, I am leaving out a lot of the design process for components on the rover I did not take complete ownership of, despite working on them. These can be found in our presentations and reports. I would love to talk about what other designs we iterated for this rover, so feel free to ask! This leaves me to discuss what we did with our rover once the designs were complete and once the manual and instructional video were complete (See Testing Phase).
​
Linked below are snippets of the documentation we created through the year in the form of presentations, reports, and the assembly manual itself, redacted as per request by our advisors at AREN and GLOBE.
Design:
It's worth noting that throughout the year, we were in close correspondance with our clients at AREN and GLOBE (NASA), and in even closer communication with our Senior Design advisors Dr. Anthony Linn and Dr. Caleb Farny. ​
Day One:
On the first day, we received a box of sensors and 3D printed parts, a shoddy parts list, and a packet of how to piece this rover together. At first glance we realized that their manual needed a rework because it was confusing and seemed thrown together at the last minute. Details were overlooked, and a plethora of teaching moments, especially for less experienced engineers (like high school students!), were glazed over.
​
We succesfully built the rover, which involved soldering and wiring the sensor package (Urban Heat Island Instrumentation Package or UHIP), built off an Arduino Leonardo, and the vehicle components made from 3D printed trusses, wheels, and motors. Through this intial build we realized a number of key points where there could be improvement in design, and more improvements as it pertained to the manual. These realizations from Day One would pave the path for our work in the coming months until the end of the year.
Our Findings:
After our intial build we reconvened with Andy (AREN) and Geoff (GLOBE) to talk about what kinds of directions we could take with our findings.
(1) Knowing that most schools don't have the infrastructure for or expertise to solder, we pitched the idea of eliminating all soldering from the build. This would be done by using wires with connectors. Even with our undergraduate understanding of wiring and soldering, their original instructions were unclear and had no explanation for why certain wires would go to certain pads. It was a stretch for an even younger student to try to understand these electrical engineering concepts, let alone even try to solder with no experience. After some debate, the decision to eliminate soldering from the build where possible was solidified. There would be no soldering when it came to the vehicle assembly, and soldering related to the UHIP would be done in-house.
​
(2) Even with no soldering as it pertained to the vehicle assembly, wires were still a mess. There are six motors with two wires each, connecting to a speed controller, connected to the RC Receiver. Without a careful eye, it can get confusing quickly. We designed a wiring plate that would press fit all the wires from the speed controller, so that it could serve as an organizational aid for assembly. This wiring channels would come as a plate attachment for the top of the RC Truss (platform containing all components related to vehicle movement, minus the motors). Piggybacking the idea of this being an organizational tool, it would also make it easier to showcase the wires in the assembly manual, and convey information about what wires were going where and why.
​
(3) The last design changes to the rover pertained to safety and weather protection. These would be press-fit motor caps and a thermoformed plastic cover over the RC Truss. Previously, the rover was susceptible to weather, especially rain, so with these adjustments, it would be able to drive outside with less concern.
​
(4) With these findings, we also came up with the idea to demo our designs and manual to a classroom of children to see how receptive they were to our manual, our improvements to the rover, and the project as a whole because again, this would be a kit for high school aged students to build and operate. This classroom demo would take place towards the end of our project.
Wiring Plate:
Along with being team lead, I choose this design to take on. Not only did I conceptualize it from the beginning, but also of the improvements to make, it seemed like I could learn the most from taking on a design like this, which involved using CAD and 3D printing, both of which are skills I love to exercise and want to continue honing my skills at.
​
The first step to creating this wiring plate was to ensure that two wires at a time could be press fit into its channels. This took a small amount of trial and error, basing the intiial print off of the diameter of the wires we would be using. The pathing for the channels was from the speed controller out to all six motors.
​
FDM PLA Prints are often varied in these tolerances so to get this right, there had to be multiple tests with certain channel shapes and diameters. Now, because these are all 3D printed prototypes, and not injection molded final products, the tolerances could be a little more lenient. My goal was to make these channels snug enough to pass as an intuitive press fit, while also not being too difficult to press wires in for ease of use.
​
Once the wires were set, the plate would go through multiple interations to improve ease of use during the rover assembly.
Wiring Plate, Version 1:
Fig. 3. Wiring Plate V1
Seen above is the first iteration of the wiring plate, meant to lay completely over the truss, under all the RC components. The notches near the middle of the plate were for wires to be able to slot into the channels, where they would be directed up, down, and laterally to all six motors. The square holes lining the top and bottom of the plate were to mount the plate to the truss, which already had square holes for other mounting purposes.
​
This iteratiion proved fruitful, but there were changes to be made for ease of use, namely, the notches were too cumbersome to try to slot wires in, so either they needed to be larger, or an alternative method needed to be found.
Wiring Plate, Version 2:
Fig. 4. Version 2 of the Wiring Plate.
Fig. 5. Picture of the Wiring Plate during assembly.
This wiring plate opted to cut the middle out completely, allowing room for the excess bundle of wires coming from the speed controller. This cutout also allowed the speed controller to sit flush to the truss. It should also be noted as well that the bend angles of the channels from the first iteration were fine as they were, so these were also applied globally where applicable.
Wiring Plate, Version 3, Final:
The last iteration was a quick one, but the channel shape was slightly adjusted, and the thickness of the plate was increased to prevent cracking, especially in the far corners. The final iteration is seen on the rover above in Fig. 5.
Fig. 5. Version 2
Fig. 6. Version 3
Testing Phase:
We discussed with our correspondents at GLOBE and AREN, and brought the idea that would could shop this kit to different schools and after-school STEM programs to gauge interest in the TerraRover and for us to test the effectiveness of our manual and video.
​
With months of communcation with experts in STEM education, professors, and teachers in the Greater Boston Area, we were able to mold our manual and video into something geared towards Next Generation Science Standards (NGSS) education. Also through our discussions we were able to secure some time with an after school program and test the manual and video on a group of high school students.
After School Program
For the sake of confidentiality, I've left the name of the after school program out.
​
We had a 2 hour time slot to evaluate the kids in their assembly of the rover, given only our manual and our video. Note that this test was only performed to showcase the vehicle components of the rover. We did teach them about the Urban Heat Island Effect and brief them on this rover's larger purpose, but this demo was only for the vehicle.
​
To standardize our findings, we drafted evaluation sheets for before and after the build, linked below is a filled-out survey:
Along with creating a survey, we were taking notes through the whole process, paying careful attention to which parts of the build were intuitive, and which parts could be more intuitive.
Testing, Takeaways:
The biggest and most general flaw with our manual was that there were simply too many words.
​
In creating this manual from scratch, we chose to be highly detailed and informative. Its intended purpose was for the reader to understand more deeply, and complete the build with a stronger understanding of this build. This is what I wanted from the initial build, but we failed to realize that few people actually want to sit down and read in this situation - they want to just build. What actually happened was that text was deterring the kids from reading at all.
​
Our manual was created with LEGO manuals as heavy inspiration. Their manuals have no words at all. Our manual has intricate diagrams AND words, so as to cater to a more intensive build involving electronics and mechanical fasteners. Our diagrams stand alone, and the words should supplement. But we realized there were simply too many words.
​
To maximize learning from this manual, we would need to make the text more approachable. Definitely more concise, but also try to frame it in scenarios that they might get stuck at. Being able to predict where someone might get stuck, and offering "extra help", is a great way to get someone to actually read the text because they NEED to -- though again, not paragraph, but more like small blurbs / explanations.
​
These suggestions only touch the surface of what improvements could be made -- there is a whole world of graphic design that caters to readability and approachability for readers. Lots of psychology and empirical research. Something in this realm would make this manual more readable for kids.
Fig. 7. Students working on the rover.
Fig. 8. Students working on the rover in the early stages of the build.
Fig. 9. Full build of Rover