AI systems and products are becoming more and more ubiquitous thanks to advances in machine learning and big data. Obtaining the right data can often be the most challenging obstacle in creating a new AI system for analyzing the sounds in a home environment.

​

Our team, Echoes, took a game with a purpose (GWAP) approach to solving this problem by gamifying audio collection. We used the Twitch platform to collect sounds in homes including but not limited to dishwasher running, coughing, footsteps and security alarm. I am the producer and a programmer in this team.

​

During the beginning of this project, we faced a huge design obstacle of gamifying audio collection. The question of how to make audio recording fun for people has never been solved before and having the constraint of using Twitch as our platform adds on to our challenges. We were able to overcome that challenge by communicating with other designers, playtesters and our client. 

​

As a producer in this team, my tasks were to have everyone in the team be clear about our goals, motivate them to move forward, assign tasks and manage our schedule. This proved to be a personal challenge for me as I was new to being a producer and found it hard to manage my time between being a producer and a programmer. I soon found out that being a producer is a time consuming role and decided to take on less programming tasks. I met with experienced producers to understand more about the tasks and how to be a better producer.

 

Though I had decided to take on less programming tasks, this became difficult as more programming tasks were added on to the backlog and I started working on the mobile application (using Unity) for recording/uploading audio files.

​

The link to the project website is provided below:

http://www.etc.cmu.edu/projects/echoes/

​