-
Gameplay screenshot
-
Save and Share screen
-
Catastrophe during testing
Inspiration
Right before the hackathon we were brainstorming ideas but eventually we got bored. And so, a laptop gets pulled out, new Youtube tab is opened, and we start watching a 2 hour long Full Game Walkthrough of WarioWare: Get it Together.
We discovered that WarioWare: Get it Together is not just one game but rather an assortment of minigames. With barely a dozen hours total to complete our project, we realized something of similar scope to a single WarioWare game was our best shot.
After watching “ WarioWare: Get it Together - Full Game Walkthrough “ for 10 minutes, one of our teammates pulled out their phone and one app caught our eyes. The app was PhyPhox, a tool we used in our high school physics classes to utilize the variety of sensors available on smartphones.
With our WarioWare inspiration, we knew what we were going to make. A drawing game that uses your phone as your way of interaction.
What it does
With just a phone accelerometer and the power gravity, we determine the tilt of a phone and use this data to control a digital drawing game. The game gives you a prompt and the objective is to get the highest rating possible by accurately depicting the required object within a minute. Of course, the challenge comes from the controls, with the only way to move the pen being your phone’s tilt.
How we built it
Our application has two main components, the (locally hosted) webpage and the backend connection to PhyPhox.
We stream accelerometer data from a smartphone in real-time over the local network. The raw json data is accessed by the computer and processed into velocity vectors and displayed on the webpage.
On the webpage, your drawing is assessed by Gemini 2.5-Flash and given a score.
Challenges we ran into
The main issue we encountered during development was the accuracy of the controls from the tilt. While PhyPhox is able to collect sensor data and report it to the program, the actual usage of this data was difficult since there is a lot of random noise and sensor drift when running the program. The acceleration based data meant that to obtain any useful information, there had to be many imprecise calculations and any slight disturbance could significantly affect the control.
To try and combat this inaccuracy, our team went through several solutions before finally settling on a system using calculated velocity data, which reduces the calculations needed to a bare minimum and prevent error, and filters to reduce noise levels that could potentially skew the data.
What we learned
Almost everything about this project was a learning experience. They say ‘failing is a learning experience,’ and as such, we failed a lot. But in the end, we were able to adapt and overcome to reach an acceptable final product.
Some highlights are: learning the PhyPhox API, the values of strong project planning, and adaptability.
What's next for Wand Drawing
The limited time scale of this project meant that our team couldn’t really expand on some of the other ideas we had for the game. Since PhyPhox has many other measurement options, we wanted to create alternate control schemes where the audio input, color from camera capture, or even the spatial positioning of the phone could be used to draw instead. Another interesting feature we were considering was a multiplayer connection, as the program allows multiple phones to simultaneously send data, and create some cooperative or competitive games using the sensor data.
Log in or sign up for Devpost to join the conversation.