
Engineering Design Workshop 2026
Friday, 31 July 2026
Day 20- Discussion and Reflection
Engineering Design Workshop- Discussion and Reflections
The end of the four-week summer engineering program provides a valuable opportunity to gather student feedback and reflect on the project ideation process. The instructors use this session as a testing platform to improve future iterations of the course. A key discussion point was the timing of team formation and project ideation. Some participants felt that forming teams earlier allowed members to shape the project together, while others suggested that incorporating more inspiring activities early on—such as working with telegraphs or weather balloons—could help generate better ideas before groups are finalized.
Integrating optional technical workshops throughout the program was highly recommended by the students to build essential engineering skills. Participants appreciated voluntary sessions that taught basic skills like wiring, soldering, and reading schematics. Many agreed that offering these workshops once a week during the active project phase makes the learning process more relevant and satisfying, as students can directly apply new knowledge to solve immediate problems. Additionally, there was interest in providing pre-program resources, such as instructional videos or a box of starter parts, to help students build a foundational understanding of electronics before arriving.
See resources at https://edw2026.blogspot.com/p/tech-notes.html
While the current four-week, four-hour daily schedule is generally effective, adjusting the program's timing presents complex logistical trade-offs. Extending the daily hours could provide more uninterrupted work time, though it might lead to fatigue and require additional snack breaks. Conversely, condensing the program into three weeks could negatively impact the ability to procure specialized parts online, a luxury not available during the intensive three-day international hackathons where teams must rely solely on provided materials. Ultimately, the consensus indicated that the existing four-week structure provides a healthy balance of time pressure and flexibility.
Balancing creative freedom with structured inspiration is crucial for helping students navigate the initial design phase. While introducing specific themes like "smart cities" or wearable technology might provide helpful scaffolding for beginners, many students argued that strict themes could limit their creativity and discourage unique ideas. Instead, participants suggested alternative methods for inspiration, such as practicing design with a hypothetical client or sharing stories of past playful projects. This approach encourages out-of-the-box thinking without forcing students into restrictive categories.
The program’s hands-off mentorship approach successfully fosters independent problem-solving and embraces the concept of failing fast. Instructors deliberately avoided micromanaging, allowing students the freedom to experiment and sometimes fail. While some students noted that earlier intervention could prevent them from wasting weeks on unviable ideas, most appreciated the autonomy to discover these lessons themselves. The primary goal remains for teams to achieve a Minimum Viable Product in a short timeframe, ensuring that the learning process takes precedence over a perfectly polished final presentation.
Expanding the program’s impact globally requires empowering students as agents of change while building strong partnerships with local educators. Instructors expressed a strong desire to help students launch similar engineering clubs and initiatives in their home countries. To achieve this, students requested access to lesson plans, structured activities, and continued mentorship via online platforms like WhatsApp to maintain a global community. Furthermore, bringing local teachers into the process is essential, as their support and understanding of the program's goals are necessary to reform traditional educational systems and foster a lasting environment of creative learning.
See examples at:
MIT Toy Design 2.00B-Electronics
Beaverworks Create Assitive Tech Online Course:
https://beaver-works-assistive-tech.mit.edu/create-challenge/create-course
Electronics Basics
https://www.engineeringdesignworkshop.com/electronics-basics.html
Arduino Workshop
https://sites.google.com/view/arduino-one/arduino-project-book
Microbit Workshop
https://sites.google.com/view/microbitaceraworkshop/microbit-computing
Thursday, 30 July 2026
Day 19- Student Presentations.
Team SMUR Radio
Our team, Smur Super Mega Ultra Requite Costa Mario, designed custom devices that use Bluetooth to send text messages and music. We initially intended to use traditional radio waves, but we pivoted to Bluetooth because it was more manageable for our timeframe. The resulting devices, named Rayoverta and Rayoberto, along with our early prototype Mauricio, can communicate up to 35 meters apart as long as there are no walls obstructing the signal. We chose to focus on telecommunications because it is the foundation of society, and we wanted to create a special way to share media without relying on Wi-Fi. Users can send text messages via Morse code, play music, and even "Rickroll" their friends completely off the grid.
We divided the technical development of our project into three parallel tracks covering the hardware, the software, and the physical design. The internal electronics rely on an ESP32 microcontroller, paired with an amplifier and a speaker to ensure the audio is sufficiently loud. The user interface features a switch to toggle between music and text modes, as well as three buttons: one for a Morse code dot, one for a line, and one to send the message. We programmed a database that translates these Morse code inputs into the alphabet, numbers, and a selection of songs. To make the design more functional and improve the acoustics, we housed the components in a sealed box that naturally amplifies the sound.
To make the radios more engaging and personable, we programmed digital faces and incorporated physical movement. Using a downloaded graphics library, we drew basic shapes, like circles for eyes, directly on the device's screen. We then created animations by rapidly cycling through a sequence of different pictures. Beyond the digital display, we did not want the radios to be entirely stationary, so we designed them to dance along to the music they play. We also built a separate spider-like prototype equipped with motors that can physically move forward, though we lacked the time to implement full directional control.
During our demonstration, we showcased how these devices function as a healthy, distraction-free alternative to modern smartphones. By inputting five dots in Morse code, for example, a user can select and play a track from our database of ten songs, a feature we included because we believe music is an essential part of life. Using this device to send texts provides a healthier communication method that removes the constant distractions of phone applications like TikTok and Reels. The radios are also highly practical for remote areas, such as high altitudes where internet and satellite connections are unavailable, because they connect directly to one another. While our current setup focuses on two primary devices, the system could easily be expanded to include more units by adding additional ESP32 microcontrollers or by upgrading to a radio module with an antenna to increase the overall range.
Team Robo Dog
Our team of students came together at MIT during the month of July to develop "Project Robo Dog." The group consists of Ariana from New Jersey, Agustia from Kentucky, Maha from Italy, Laya from Barcelona, and Erica from Cambridge. We all joined this program because of our shared interests in various fields of engineering, including biomedical, electrical, mechanical, and industrial applications.
Although we initially intended to design a prosthetic leg for a three-legged dog, the logistical challenges of testing led us to pivot to a fully motorized robot dog. We had briefly considered combining prosthetic front legs with motorized back legs, but ultimately decided a complete robot was the most feasible option. Our primary motivation was to integrate mechanics and electronics to help injured animals, especially after our research revealed that many existing dog prosthetics are essentially just pipes or metal extensions lacking functional joints. By heavily researching a dog's natural walking gait, we gained valuable mechanical knowledge that could theoretically be applied to future, non-motorized prosthetic structures.
To arrive at our final mechanism, we iteratively designed and built at least six or seven different prototypes. The earliest version utilized gears but did not closely resemble a dog, while a second prototype focused on knee movement using a motor and an elastic band. Subsequent iterations explored using a single motor for both the hip and knee joints, as well as a four-bar linkage system initially designed to move all four legs with only two motors. We eventually combined the four-bar linkage concept with inspiration from existing cyberdog models, settling on a system that uses four gears and two motors per leg to control both the hip and knee movements.
Building our final product, affectionately named "Carlita," required a combination of laser cutting, computer-aided design (CAD), and custom electronic software. We used CAD to design the entire dog and its gears, relying on trial and error to perfect the spacing and alignment. For the physical build, we laser-cut wood and acrylic materials to maximize the structure's overall strength and stability. To make Carlita move, we painted her with fun rainbow Dalmatian spots and powered her using an ESP32 microcontroller connected via Bluetooth. This electronic system controls a total of eight motors—two for each leg—which are carefully synchronized through our custom software to simulate a dog's natural walking motion.
Despite encountering numerous challenges throughout the prototyping phase, we successfully assembled a fully functional, four-legged robot in our final week. Building a single working leg took a considerable amount of time, making the final assembly a significant achievement. This success would not have been possible without the guidance of our mentors, Jordina and Anna, as well as the support of the program organizers, including Chris, Gill, and David.
Our first version of the prosthetic leg was designed to improve upon the attachment methods and functionality of current models. Existing prosthetics often consist of a simple straight bar that attaches to the dog, but our intention was to incorporate a functional knee joint to better mimic a real leg and assist the dog's natural movement. Furthermore, we intend to use materials like silicone to capture the exact shape of the dog's residual limb. This custom molding technique would allow us to create highly specialized, comfortable prosthetics tailored to the individual animal, rather than relying on a generalized model.
The electronic control system also allows us to fully customize the dog's walking gaits and drive modes. We have complete programmatic control over exactly what the robotic legs do. Depending on the desired movement, we can set the legs to operate in a synchronous pattern, an alternating sequence, or whatever configuration is needed.
While the current prototype is not perfectly stable, achieving independent robotic balance was not the primary focus of our research. The current instability is largely due to the trial-and-error approach we took with the gear tolerances, which resulted in slightly shifted parts rather than precisely machined components. However, our main objective was to research how to engineer a functional prosthetic joint. In a real-world application, a dog fitted with this prosthetic would still have two or three of its natural limbs intact to provide natural balance, meaning the prosthetic only needs to successfully replicate the mechanics of the missing limb.
Team Starglasses:
The team introduced their project, Starglasses, along with their international members and the presentation agenda. The team consisted of Zanny from Mexico, Mara from Turkey, Marzara from the United States, and Victoria from Italy. They were supported by mentors from Spain, Turkey, and Italy. The presentation covered the origins, initial ideas, iterations, problem-solving processes, technical aspects, and the final product of their Starglasses project.
Starglasses is a wearable device that utilizes an optical system and a data library to display astronomical information. The glasses reflect a multitude of stars, constellation names, and planet names. Additionally, the display shows the distances to these celestial bodies.
The inspiration for Starglasses stemmed from a desire to combat Boston's heavy light pollution and recreate the clear night skies of the Italian countryside. One team member noted the stark contrast between the bright city lights of Boston and the starry skies visible from their grandparents' rural home in Italy. They envisioned an augmented reality device, similar to Meta AI glasses, that would allow users to look around and see real stars and planets projected inside the lenses, essentially bringing a clear night sky to the city.
Understanding and implementing the optical system proved to be one of the most difficult challenges the team faced. The system works by projecting a display screen onto a beam splitter, which then reflects the image into the user's eyes. This setup allows the wearer to see both the real world and the projected screen simultaneously. With the help of their instructor, Jim Bales, the team tested various lenses and focal lengths to better understand the optics.
The physical design of the glasses required significant iteration due to time constraints and unwieldy early prototypes. Initially, the team planned a complex optical system using an LCD display, a convex lens, a combiner, and a beam splitter, which resulted in a massive frame playfully dubbed "Frodo's nose crusher" because of how uncomfortable it was. Realizing the four-week timeframe was too short to build everything from scratch, they simplified the design to use only a beam splitter. This adjustment eliminated the long focal length problem, and they utilized a laser cutter to create a more manageable frame.
Technical difficulties with outdated hardware forced the team to implement a successful backup plan using a transparent OLED display. The original plan involved connecting a small, older projector to a Raspberry Pi, but the projector arrived just three days before the deadline and the connections were too difficult to manage. As a fallback, they used a small, transparent OLED display and a mirror, which successfully allowed users to see the simulated stars. The final program utilizes an IMU sensor to track head movements, changing the projector's view as the user looks around.
To save physical space on the device, the team abandoned a hardware GPS in favor of hardcoding the geographic coordinates for Boston. Rather than using a bulky GPS module to track position and time, the glasses calculate star placements based on the static coordinates of MIT. The system pulls the current time from a small internal battery and automatically calculates the correct positions of the real-world stars and constellations every time the program boots up.
The prototyping phase involved creating numerous models out of various materials to test designs and optical components. Alongside the painful "Frodo's nose crusher," the team built laser-cut "party glasses" to test basic designs and different displays. They also created approximately ten different cardboard prototypes, referred to as "Men in Black sunglasses," and experimented with 3D-printed casual frames. Detailed note-taking by a highly organized team member helped keep this rapid four-week prototyping phase on track.
The presentation concluded with a final demonstration of the working glasses, a Q&A clarifying that the digital constellations accurately reflect real stars, and heartfelt thanks to the staff. The team confirmed that their program displays true star positions along with connecting lines and text. They then expressed gratitude to their mentors, including Nade, Ari, Natte, Sheila, and Juliano, who helped with 3D printing and laser cutting. They also thanked instructors and staff members Jim Bales, Jonathan, Ed, Kim, and Adam for their invaluable support throughout the project.
Team Angel Wings:
The team developed a prototype bird wing exoskeleton designed to assist parrots with flight. The project, affectionately modeled on a dummy parrot named Harold, is targeted at birds suffering from wing defects or improperly healed bones. To ensure the device is completely removable and harmless, the prosthetic is mounted on the bird's back via a hand-sewn vest. Clear strings attach to Velcro near the bird's joints, avoiding any direct or permanent attachment to the animal's body.
Achieving symmetrical wing flapping required significant troubleshooting and multiple mechanical iterations. Initially, the team designed and 3D-printed a diamond mechanism on a track so that pulling one side would symmetrically push the other. However, excessive friction rendered these early prototypes ineffective, prompting a shift to a lever arm mechanism. This final mechanism eliminates the friction issue by using attached rubber bands that snap the levers back into place.
The physical model of Harold the parrot evolved from basic materials to a realistic, weighted display. Early iterations included a simple cardboard and fabric model, followed by a skeleton constructed from laser-cut wood to resemble a bird's body. To make the model look like a real bird, the team covered the wooden frame with paper mache, painted it, and added decorative feathers. When they discovered the completed model kept falling forward onto its face, they successfully counterbalanced it by stuffing metal balls into its lower abdomen.
Although the current model relies on a manual string mechanism, the team also made significant progress toward motorizing the flapping motion. Because a real parrot cannot lift a defective wing using the force of its healthy one, the team experimented with Lego gears and an Arduino-powered small motor. They used Motion Gen simulation software to translate the motor's circular motion into a realistic wing flap, though they ultimately needed more time to successfully mount and finalize the motorized system onto the working prototype.
The project was motivated by a gap in modern avian veterinary medicine and brought to life using a variety of digital and physical fabrication tools. The students noted that while prosthetics for dogs and humans are common, miniaturizing these devices for birds remains incredibly difficult, leaving a severe lack of similar options for parrots. To step up to this challenge, they utilized the 3D printers and laser cutters at the Edgewater Center as their primary fabrication tools. Additionally, they relied on Onshape and Fusion 360 for 3D modeling, and the Arduino IDE for software development.
Looking ahead, the team hopes to test and refine their design on a live bird to evaluate its overall comfort and functionality. While they are currently unable to test on a real animal, their future plans involve fitting the vest and mechanism on a bird to identify and fix any emerging problems. Furthermore, they plan to redesign the physical shape of the artificial wing to act as a protective armor for the bird's natural wing, ensuring the device does not cause any unintended damage or chafing
Team Glider Goats
The presentation began with the team outlining their personal and project goals, as well as the initial design for their remote-controlled drop vehicle. The students shared that their primary personal objectives were to gain hands-on building experience, create something interesting, and learn more about CAD software. Their overarching project goal was to build a payload device that one person could drop while another steered it from the ground using a remote control. To begin testing this, their first electronic prototype utilized an Arduino Nano mounted on a breadboard and connected to two servo motors.
Although the team originally intended to launch a traditional rocket, they quickly realized they needed to adjust their scope for the short four-week timeframe. Initially, they sketched a system featuring an adjustable parachute, springs, and a central servo motor, but they concluded that launching a real rocket would require ordering complex parts and designing intricate systems that would take too much time. They used Onshape to design their CAD models and soon shifted their focus toward a dropping mechanism rather than a launching one. Their very first physical prototype was made of a PVC pipe wrapped in foam and duct tape, weighed down with a rock at the bottom, and equipped with a simple cloth parachute. This early model allowed them to observe the falling dynamics, specifically testing if adding weight to the bottom would help the device auto-correct itself and land upright.
After a few days of experimentation, the group established three definitive goals to ensure their project was both realistic and aligned with their original vision. First, they wanted the device to consistently land upright instead of falling on its side. Second, they aimed to effectively control the steering servos using their programmed Arduino code. Finally, they prioritized a soft landing by incorporating a foam plate to prevent the device from breaking upon impact. To support the landing, they went through several iterations of deploying legs. Their first leg design was a heavy 3D-printed part, which they quickly discarded in favor of a lighter wooden version; however, the wood was too fragile and repeatedly snapped on impact. Ultimately, they settled on a spring-based suspension system, similar to those found in cars and bicycles, to absorb the landing shock.
To further improve their controlled descent and ensure a soft landing, the team finalized their electronics and replaced their standard parachute with a parasail. They swapped out their large servo motors for smaller ones to reduce the overall weight, which directly supported their goal of a gentler touchdown. The final paraglider design featured two servo motors attached to pulleys. By using the remote control, the user on the ground could rotate the pulleys to shorten the string on one side of the paraglider, successfully steering the falling device. Despite sustaining damage during numerous test drops—including one drop from the building's fourth floor where it hit a tree, did a front flip, and still safely landed—the final product proved functional. The project, originally inspired by real-life aerial payload delivery systems, concluded with the team thanking their mentors and support staff for their guidance.
Team "Grab and Go"
The "Grab and Go" robotics team drew inspiration from several real-world machines and nature to design their project. The team based their mobile robotic car on the mechanics of trash trucks. For the robotic arm, they looked to industrial robots, toy robots, and robotics competitions, while the grabbing claw was specifically modeled after the anatomy of bird claws. During their presentation, they showcased videos and photos of the robot in action, including test drives and picking up lollipops, highlighting their use of tools like Onshape, Fusion 360, and the micro:bit.
The physical design of the robot's arm and claw underwent several iterations to improve functionality. Initially, the team constructed a three-fingered wooden prototype for the claw. However, after encountering issues with the gears, they transitioned to a two-claw design, which was ultimately 3D printed for the final version. The robotic arm also saw revisions; the first prototype functioned, but the joint where the claw attached was inefficient, prompting the team to completely redesign a new component using CAD software.
To support the arm and enhance user control, the team had to build a custom car chassis and a glove-based controller. Originally, they intended to use a pre-built toy car that could move sideways, but they accidentally destroyed it while attempting to swap out the motors. Consequently, they built a new car from scratch using more powerful motors capable of supporting the robot's heavier weight. For operation, they upgraded from a basic button-push controller to an innovative glove-based model, allowing the user to control the robot by tilting their hand and using built-in buttons and potentiometers.
The students faced and overcame several technical hurdles regarding coding and electronics integration. At one point, a software bug caused the robotic arm to randomly move and "dance" on its own, which they successfully corrected through the code. Additionally, a simple syntax error—an accidental apostrophe typed instead of an enter keystroke—cost the team two entire days of debugging. To organize the physical hardware, they mounted their components onto an acrylic sheet and used yarn to manage the wires, connecting everything to a single micro:bit.
Although presented with a humorous demonstration of grabbing dolls to deal with "annoying" mentors, the robot was ultimately designed with a serious assistive purpose. The team joked that they built the robot to grab mentors who constantly changed the project parameters or ran away when asked questions. To demonstrate, they invited a friend to help drive the car and pass out lollipops to the front row of the audience. However, they clarified that the true purpose of the "Grab and Go" robot is to assist individuals with mobility issues, such as those confined to a bed due to injury, and to help people manage emotional problems.
Day 19-Presentation Preparations

- Mentors should come prepared with better tech skills, particularly in the areas of electronics, microntrollers, and coding
- Students need to be more adept in teaching themselves technical skills- in the workforce, professionals are expected to pick up new skills on their own
- After the initial ideation, students should research precedents- how have others solved this problem- and document websites and projects examined
- Presentations should include:
- Project Title and Team Members
- Project overview- what is the problem they are trying to solve?
- Precedents
- Bill of Materials
- Technical Diagram
- Circuit Diagram
- Code and design files
- Results
- Next Steps
Wednesday, 29 July 2026
Day 18- Last Day for Project Work!
Team One Wing Angel- Practice Presentation:
Chris strongly urges the team to take a much-needed break to ensure their own well-being and to set a positive example for the students.
With only one day left until their hard work pays off, Chris also emphasizes the critical need for ongoing self-care and hydration.










































