Carnegie Mellon ECE Capstone, Fall 2025: Songting W, Frances A, Alex K
This week, we all worked on polishing our game up in time for the final demo. Frances worked on attaching the hardware to the frisbee itself, Michael worked on tweaking the values to make the game playable with the frisbee controller, and Alex worked on helping debug the implementation of the hardware and software.
We believe we are on track for our Demo, and we are having a lot of success with the implementation of both, hopefully creating a fun experience for all of our users!
This week, I worked on finally getting the hardware off of the breadboard and into our casing. I required all of our connections using female-to-female jumper wires, and attached our battery which finally arrived. I also prototyped 3 versions of our hardware casing out of foam-core, adding an openable flap, so we were able to reset the Arduino and disconnect the battery. The hardware casing also included an IMU pocket, so that it wouldn’t jostle around and potentially skew the data while the users were playing the game.
We are on track to have a successful demo. This upcoming week, we will work on the final report and video.
This week I worked on creating the final slideshow and practiced what I was going to say for the presentation. I believe it went pretty well and I was able to accurately cover what I wanted to say. After Monday, I moved on to helping the group as a whole complete the remaining parts of the project. This includes fully implementing the CV, helping debug and giving ideas to Michael for the software part of the game, testing the controller and helping build the final hardware design.
I am very excited to show off our completed project on Monday. The project will be completed on time and it will be extremely fun. I hope everyone enjoys playing our game!
This week, I focused heavily on integrating our hardware, CV module, and software components. I added new gameplay elements and, throughout the integration process, identified and resolved several bugs, connection issues, and other inconsistencies. I also worked on improving the overall player experience. For example, smoothing camera motion, refining the camera-follow behavior, and removing unnecessary obstacles from the map.
Overall, our team is on track for Monday’s demo, and we feel confident that the experience will be both engaging and meaningful for the audience. After the demo, I plan to work on the final report and begin collecting accuracy metrics and user experience feedback.
This week we were successfully able to wirelessly communicate data between the Arduino and the computer. We are also able to successfully power the Arduino with a battery. This week, Michael further developed the game experience by making a scoring system with targets that the users can hit, and we have also decided to no longer use a ToF sensor, and instead use the CV model to measure the distance the frisbee is off of the ground. We were able to attach the wrist strap to the frisbee controller, however it was too flimsy, and broke, so we have put in another order for another wrist strap. Next week, we are planning to get the 3D printed casing attached to the frisbee with the hardware inside. We are also planning to finalize the communication protocol between the CV computer and the Unity Game computer in hopes to finally implement the CV data.
We are still on track to finish in time.
For this week, we spent a majority of time setting up for our interim demo. Alex and Frances worked on getting a wireless communication working between the controller and the Unity Game, as that is a crucial and critical component of the product. Michael added more parameters to the frisbee’s path of flight within the Unity Game, and fine tuned the game’s physics. Next week, we are planning to finish setting up the wireless communication, and set up the communication from Alex’s computer (which hosts the CV) to Michael’s computer (which hosts the Unity Game). We are mostly on track to finish in time.
This week, we did a lot more development into the implementation of our Design. Frances worked on figuring out how to get the ToF sensor working, as the IMU was transmitting data successfully, and we needed to also get the ToF sensor working as well. Michael worked on developing the game environment and smoothing out the physics of the model. Alex found a better CV model with a lower latency for our design.
We are fairly on track, however our Gantt chart had to be updated to account for using CV as well as giving ourself more time to get the hardware implemented. We are in a good spot for the interim demo, as we successfully are able to connect the hardware movement with a software response.
This week, we have made the decision to not use the ToF sensor to measure distance off of the ground, and instead decided to measure distance using the CV system we already had set up. This was due to a number of reasons; firstly, we were quite concerned about hitting our weight requirement, as the Arduino Nano, IMU, casing, battery, wires, and peripherals were quite heavy. Secondly, having to connect the ToF sensor as well as the IMU to the microcontroller might have required the use of a breadboard or other peripherals we weren’t sure we had access to. Since we already had the CV system set up, it was quite easy to implement distance off of the ground, since the CV was already measuring the user’s hand position, relative to their body and height.
This week, I spend a lot of time trying to debug the ToF sensor and get it working via I2C protocol, which ended up not being necessary for our project. I was also able to start looking into 3D printing files for our casing around the whole frisbee.
This week, I spend a lot of time figuring out how to make the hardware wireless. We had previously attempted to communication with the Arduino Nano through CMU’s Device WiFi, however this had proven to be especially complicated, since it was difficult to connect the computer to CMU-Device, and both the computer and the microcontroller needed to be on the same network in order to communicate.
As a result, we have successfully implemented a WiFi communication using hotspot, where both the devices are connected to the hotspot. This was a very critical step in our project, and I verified the correct communication by powering the microcontroller with a different power source.
For this week, I spend a majority of time trying to wire the ToF sensor. I was having lot of compilation issues, where the Arduino IDE imported libraries weren’t fully supported with the ToF sensor we specifically chose.
After a bit of tinkering, I found that the specific sensor we purchased was unable to communicate via SPI to the Arduino Nano, although SPI is supported on other models of ESP32 microcontrollers. As an alternative, we are able to use I2C to wire up the ToF sensor. The drawback with using I2C is a higher latency, however the distance the frisbee is off of the ground does not need to have a high sampling rate, therefore making I2C a suitable communication protocol in our use-case.