Chip Chatter

Hi, my name is Adele and my name is Abigail, and we both are members of C2S2's Project Management subteam. In our last post, we assisted the System Architecture team in creating a system diagram of what the team's proposed proof of concept chip, tailored to our campus partner, would look like. There still are a few moving pieces with projects, but the system diagram was updated to include the Software subteam's classifier, which determines if a bird call is a Scrub Jay or not, and System Architecture's particle board as a method of offloading data. The most updated diagram was presented to the team during the final presentation towards the end of the Fall 2023 semester by System Architecture. With projects for each subteam now catered specifically towards the campus partner and the team preparing for taping out chips this semester, one of Project Management's primary focuses has transitioned back to collaborating with our campus partner. Additionally, we wanted to ensure that he was equally as informed of the team's efforts so that we could all be on the same page and address concerns or potential issues early on during the design process.

To better facilitate the collaboration between C2S2 and the campus partner, we primarily worked on updating our campus partner on the team's progress and the system diagram. To effectively communicate a high-level overview and get more members of Project Management involved with sharing the technical milestones and works of the team, each member of Project Management spent time sitting in different subteam meetings and talking with members about their projects to learn about recent progress. Additionally, we each met with Abigail and Akshati to learn more details about the system diagram to have a thorough understanding of the inner workings of the chip and to be able to discuss further modifications with the campus partner.

During our meeting with the campus partner, we were not only able to explain some of the key features of our system diagram, but also discuss and address some of his concerns. One of the main concerns we discussed was more about the team's classifier and the idea of having different modes of functionality to address the challenge of limited battery life. Although this is still a work in progress for the team, one of the team's proposed solutions to lengthen the deployment period of a tag was to implement various modes of activity for the tag such as a sleep mode in addition to the traditional on and off mode. In this sleep mode, the tag would remain on but be reduced in functionality. More specifically, the tag would not be recording all audio data and instead would remain on standby waiting for the right conditions, such as the call of a Scrub Jay, determined by Digital's classifier, to turn on and start actively recording data. This was a point of concern for the campus partner because currently in his research, he listens through all of the recorded audio samples for a variety of sounds beyond the calls of a Scrub Jay to better answer what type of call they are vocalizing and determine how different calls relate to their behavior or interactions with other Scrub Jays. For example, we learned that he might listen to some noises from the environment like the wind and trees to get a better understanding of how the bird may be interacting with its environment such as flying or hopping between branches. Based on our conversations, we also learned that having the additional background noise may provide insight into what other birds it may be communicating with. Moreover, because his research in linking Scrub Jay vocal calls with their behavior and movement can be rather broad, there is a wide variety of possible sounds of interest, so it would be up to the team to decide what we thought would be interesting for our tag to switch to an alert mode to start recording data.

Although our idea of multiple modes for the tag, especially the concept of not recording all the data, may not align with the specific research objectives of our campus partner, it is in line with our collaboration agreement to develop a chip that addresses the challenges outlined by our campus partner. From these discussions, the campus partner has provided invaluable support by offering data, such as recordings collected with his current tags and sounds of interest. He also has been providing his general insight into other details we should be considering. For instance, we were able to learn and share more information on general Scrub Jay behavior, including typical activity patterns and frequency of vocalizations with the team. The discussions with the campus partner have also prompted us to explore additional criteria while developing classifiers, such as sound amplitude, to determine the tag's operational mode. This collaborative exchange of knowledge ensures that the team's chip remains informed by real-world observations and aligns with the overarching goals of our partnership.

Beyond conversations between Project Management and the campus partner to increase collaboration between the team and the Lab of Ornithology, we are happy to announce that the campus partner is finally active on Slack. This step allows any other members of the team to be able to directly contact him with quick questions, such as the technical details about his research as they are working on projects, without necessarily having to wait for a member of Project Management to coordinate a meeting. This speeds up answering any questions so that subteams can work faster toward their goals. In similar news, we also were able to coordinate having the campus partner attend an all-team meeting to present his work to the team. This was helpful for the team to be able to directly ask any questions they might have based on his current research and data findings. With this, the team also gained more insight into what is meant by "good" versus "bad" background noise as well as the particular Rattle Call that he studies. Moving forward, we still will be working on building the team's partnership with the campus partner and the Lab of Ornithology, as well as developing Project Management's skills in facilitating collaboration and communication by continuing to have these discussions with the campus partner and checking in with subteams.