Made by joyceker

To analyze how long the user has been sitting and give them ambient reminders that slowly become more persuasive to encourage them to get up.

Created: February 14th, 2016



BuzzButt started with a prompt, “ Take information in an environment that is usually not noticed or under utilized, analyze it and project it back into the world in an unobtrusive way. After lots of ideation and iterations we ended up with BuzzButt, a portable cushion that records how long you have been sitting and starts with gentle reminders to get up that get progressively more persuasive. They buzz in a pattern inspired by Beethoven’s 5th Symphony. This persuasiveness encourages the user to get up and take a break

Diagram of the sensor placement


With an increasingly sedentary lifestyle, more health problems are starting to arise. These health problems associated with sitting are not alleviated with exercise but rather with sitting less. Therefore we wanted a way to analyze when people had been sitting too long and give them gentle increasing reminders of getting up. We have it so each consecutive 30 minutes the reminder gets progressively more intense. This way the user can concentrate and ignore the buzzing when they are focused. Though the human mind can only focus for so long. The reminders get more intense until they are more persuasive encouraging people when they are less likely to focus to get up and walk around.



Round FSR

9 small vibration motors

1 large vibration motor

1 Arduino

5 10K resistors



wiring + alligator clips

1 5V power source

Seat Cushion



We first started out with the ideation phase looking through the different sensors and outputs available to brainstorm ideas. We settled on using eye tracking data and pressure sensors in the seat to tell when the user had been sitting too long and was no longer paying attention as a way to figure out when to let the user know if they had been sitting too long. This was also to encourage them to look away and take breaks from work. We originally wanted to have the output as a light strip that lit up and encouraged the user to walk away from their work station. Potentially with a small figure that had servos as their arms to direct the user similar to a flight strip. We also thought about using the sound of rushing water that progressively got louder to encourage the user to use the bathroom thus encouraging other healthy bodily functions. The issue with the two outputs was that it was only good if the user was working in a private workspace. If they were to work in a public place these outputs would disturb other people working. Additionally there would be issues if multiple people were using the same system at the same time, causing overlapping and confusing feedback. Therefore we decided to approach this issue by giving haptic feedback. 

Research in the library to see what sorts of input we could use to alert users without fully distracting them
Img 20160130 161407995.thumb

When we tried to explore the different ways of monitoring whether or not the person was no longer focused we explored the different options. With eye tracking we didn’t have the external hardware or the downloadable software. Everything we found on GitHub for open CV or TSPS that used the computer camera was in progress and not ready to be used. Creating a keylogger would have taken a long time and we fell into issue of figuring out what we would log as when the user was no longer focused. If we were to use a microphone, a person might be focused with or without typing depending on what they were doing. We also thought about incorporating a heart rate monitor to encourage the user to get up and exercise, but monitoring their starting heart rate while working and only switching off the buzzing when their heart rate had risen. This raised the issue that the heart rate monitors were not accurate enough, often not even recognizing a person’s heart beat. We though about using openTSPS to record if a person was fidgeting too much. The issue with this was the accuracy of computer vision. We set a threshold for figuring out if a person was fidgeting too much, but there needed to be a fine calibration of openTSPS to accurately measure if someone was moving or not. Even then, there was a lot of inaccuracies, and didn’t want to set the threshold to high that a person had to be rocking around on their chair for the camera to sense them moving. This issue and the fact that there was a lot of issues sending information from Processing to Arduino through serial caused us to abandon the idea due to time constraints. We also originally had a snooze button for the system that we implemented through an IR breakbeam. Which no longer worked after we implemented the ever increasing reminders to get up. We also figured out how to how control multiple different outputs. Allowing us to control the small vibration motors separately from the larger one, allowing us to create Beethoven's 5th symphony for the user's buttocks. We did all this while learning physical computing for the first time. Other difficulties we faced was fragility of the wiring on vibration motors, which kept breaking off while the vibration motors were activated. There were a lot more vibration motors, but only 10 of them survived to be implanted in the cushion. 

Luke testing out the round FSR to make sure it works
Img 20160129 165926874.thumb
Late night lab sessions
Img 20160130 202245471.thumb
Reactive Spaces project: BuzzButt
Luke Simon -
Share this Project


48-734 Reactive Spaces and Media Architecture

· 1 member

How can embedded computation change the way we experience our surroundings? What is the value of creating dynamic spaces? This class will focus on the design and prototyping of reactive spaces. Ove...more


To analyze how long the user has been sitting and give them ambient reminders that slowly become more persuasive to encourage them to get up.