SimFeel (UC Berkeley Capstone)
Technical Leadership Award Nominee
https://funginstitute.berkeley.edu/programs-centers/full-time-program/capstone-experience/
Left to right:
Colin Yang (Mechanical Engineering, Product Design)
Julia Tanner (Mechanical Engineering, Biomechanics)
Yunchen Wu (Mechanical Engineering, Control Robotics)
For my masters project, I was part of team consisting of two other mechanical engineering students to create a wearable haptic feedback system for virtual reality applications. As the team's communication and design lead, I headed the conceptualization and prototyping of the device, resulting in a functioning product to showcase on demo day.
Based on a user survey conducted by my team, it was revealed that more than 50% of current users experience difficulties with VR controllers. Moreover, over 60% of potential customers are deterred by the cost of VR, which can cost over $4,500 when incorporating haptic feedback.
To address this, we created SimFeel to replace the controllers and bring the sense of touch to VR at an affordable price. By utilizing off-the-shelf materials, such as flexible sensors, coin motors, and an Arduino, we were able to keep the cost to below $100. The device's functionality was also easy to understand for new users: while wearing SimFeel, a more dramatic bend in the fingers resulted in a stronger vibration at the tip of the fingers. This linear relationship made it seem so that the user felt something in their hands when grabbing an object in the VR world despite nothing being there in the real world. In addition to the more immersive experience for the user, we added Bluetooth capabilities and had the entire device run off a 5V battery source, making it easier to communicate with the software while being a safer and more portable device.
In addition to the market research and the user survey mentioned earlier, the team conducted two separate user tests to make the device more human-centered. The first test asked five participants to bend each finger and try different hand gestures while wearing SimFeel. This allowed us to measure the maximum readings of the flex sensors so that we could make the vibrations more comfortable and distinguishable. The second test involved a controller wearing SimFeel and a participant wearing a variation that mimicked the vibrations felt by the controller via Bluetooth. The controller bent their finger in a specific sequence, and the participant was asked to identify the order that the fingers were being bent while relying only on their sense of touch. With 13 of the 15 participants getting all three sequences correct, we demonstrated SimFeel's ability to send reliable signals as well as the potential to communicate between the virtual (controller) and real (participant) worlds.
Learn more about SimFeel:
DexPlay (Human Dexterity)
Winner of the Best Interview Narrative Award
https://edg.berkeley.edu/course/ahd/
Left to right:
Elyse Scalia (Mechanical Engineering, Product Design)
Colin Yang (Mechanical Engineering, Product Design)
Jessica Hsu (Mechanical Engineering, Control Robotics)
For the final project of the human dexterity course, teams were tasked with prototyping a novel non-invasive solution to a disability-related challenge, focusing on upper-limb mobility, strength, or comfort. Our team worked with an individual who had to give up playing the guitar because of the limitations put on his precise finger movement abilities by his essential tremors.
Our design process started with high level research, learning about the condition and currently available ways to mitigate the symptoms. We then conducted a thorough interview with the individual, identifying interpreted needs from both the interviewee's responses to our questions and the anecdotes he shared with the team. We particularly honed in on his connection to music, and how living with essential tremors had limited how much he can play his guitar. The interviewee shared that while he has tried devices that generally stabilize his hands, their bulkiness typically got in the way when having to reach around the neck of the guitar.
Using this information and multiple rounds of working with our interviewee on the design, we came up with DexPlay, a wearable device that dampens vibrations around the wrist. Highlights of our prototype include a purely mechanical design to reduce strain caused on the arm by the weight of the device, as well as having the mechanism sit entirely on the forearm so that it is less visible to the audience when playing the guitar.
Learn more about DexPlay:
Thomicory (Simulated Bass Player)
Thomas Guan (Mechanical Engineering)
Ryan Ho (Mechnical Engineering)
Miguel Montalban (Mechanical Engineering)
Colin Yang (Mechanical Engineering)
MEC 235 at UC Berkeley is a course on the design of mechanical systems, with an emphasis on microprocessors to control machine activities, acquire and analyze data, and interact with operators. Students learn how to utilize LabVIEW as a language for test and data acquisition, as well as developing an embedded system programmed in Python. Key requirements for the final project included: real time, through the crystal frequency on the ESP32; multitasking, which differs from completing multiple tasks; and creating a foolproof graphic user interface (GUI).
We designed Thomicory to simulate a live bass player. Once the user inputs a melody to play using the GUI, they are asked to swing a baton like a conductor. The ESP32 gathers accelerometer data from this baton and sends it to LabVIEW to calculate the BPM, all in real time. This continuously sets the tempo, which determines how fast the the indicators on the GUI and the physical LEDs on the hardware light up. Additionally, the team added a buzzer that plays the notes at the corresponding frequency to make the experience more realistic. Because these features are accomplished using multiple loops with varying completion times, we were able to truly achieve multitasking (which is harder than it sounds). The end result is a robotic band member that plays a user defined piece at a user defined tempo, accompanied by hardware and aGUI that show the current note being played.
Learn more about Thomicory:
Locket (Market Testing)
Thomas Guan (Mechanical Engineering)
Colin Yang (Mechanical Engineering)
Iris Yang (Materials Science and Engineering)
ENGIN 270G at UC Berkeley exposes students to a wide variety of product and marketing approaches through analysis of real-world companies, primarily focusing on how modern methodologies shape go-to-market and product development strategies.
As part of the final project for this class, my team compared the apps BeReal and Locket to come up with the value hypothesis that the number of users for the latter would be greater than the former by 50 percent. To evaluate the validity of this claim, we ran several "lean startup" methodology market tests:
App download metrics to establish where Locket currently stands compared to BeReal
Blind survey on Instagram to test which app was selected based on first impressions
Blind survey on Discord to test if app name/logo creates a bias
Focus group with couples and singles to test if different target audiences have different motivations and preferences when using the apps
A/B testing with a different group of couples and singles to test if the lack of an alternative plays a role in retention rates
Once this data was gathered, we performed a SWOT analysis, created a target user persona, and made a go-to-market plan for Locket. This included ways to build user validation through social media and partnerships, as well as strategies on how to expand into new markets.
Learn more about Locket Market Testing:
AssistMe (Human-Centered Design)
www.behance.net/gallery/158948949/AssistMe
Left to right:
Alex Sedov (Mechanical Engineering)
Thomas Guan (Mechanical Engineering)
Roy Yang (Mechanical Engineering)
Colin Yang (Mechanical Engineering)
Not pictured:
Siming Jin (Design)
ME 292C at UC Berkeley focuses on designing customer-driven engineered products by going through the full design process and applying human-centered design principles. Our team worked closely with the blind and visually impaired community to design a device that would help bridge the gap between them and sighted people when asking for help. This included in-depth interviews and observations, use of tools like user personas and weighted decision matrices, and live prototype testing. For a more comprehensive summary, I urge you to check out the team's Behance page.
Learn more about AssistMe:
Knee Brace (Robotic Locomotion)
Julia Tanner (Mechanical Engineering)
Harper Xu (Mechanical Engineering)
Roy Yang (Mechanical Engineering)
Colin Yang (Mechanical Engineering)
ME 239 at UC Berkeley is a course on how to use kinematics, dynamics, control algorithms, embedded microcomputers, and mechanical components to design artificial legs. Topics covered during the semester included the analysis of bipedal behavior (walking/running on level/inclined surfaces), various actuator technology, and hydraulic/pneumatic control systems. Students then proposed a practical application of this material in the orthotic and prosthetic industry, assessing the performance of their product in combination with the individuals wearing them.
After realizing how much stress is repeatedly applied on the patellar tendon, our team decided to work on a passive damping knee brace meant for recovery use. This started with a mathematical analysis of the ground reaction force during the gait cycle. Once the spring constant needed to support the maximum weight was found, we came up with several designs for a product, only to realize that there were limitations preventing them from working as intended. Primary concerns involved whether the leg would be able to clear the ground during the swing phase, as well as if the brace would be able to lock during the stance phase. A design proposal paper was studied to gain inspiration, but motion capture analysis done by my team revealed flaws that would render the design useless. Using this information, the team concludes by suggesting three new ideas for future applications, including the use of a gear set, a motor, and a pressure sensor.
Learn more about Knee Brace:
