MIE.24 – SPACEBOT
Team Members Heading link
- Usman Baqai
- Jackie Camarena
- Sergio Galindo
- Hope Gomez-Peck
- Michal Guzek
- Ethan Morrical
Project Description Heading link
SpaceBot is a data analytics company that produces devices to read, interpret, and store spatial data from a variety of different environments. Each device contains a Raspberry Pi computer accompanied by three sensors responsible for measuring temperature, humidity, air quality, and sound events for detection and characterization. All components are housed inside a plastic case which is mounted on a wall during operation. The data collected is uploaded to a cloud, which can be easily accessed by the customer. Each device displays real-time information about environmental conditions and any activity located in a certain space.
The operating temperature of the main processor causes a temperature increase inside the housing cavity, which disrupts the temperature and humidity readings of the device. Additionally, excessive noise and vibration from the particulate sensor disrupt the sound recognition algorithm, creating inaccurate predictions of sound events.
The end goal of this project is to engineer a new housing design for the SpaceBot device, which normalizes the temperature and humidity readings while mitigating unwanted noise and vibration from the sound detection algorithm. In doing so, SpaceBot will no longer have to program offsets and corrections in their tabular data and will not need to make any changes to the sound detection algorithm.
The group generated five different prototype designs using SolidWorks 3D modeling software, which went through a simulating and testing phase to verify them. ANSYS Fluent simulations were used to create temperature distribution maps within the housing designs, highlighting any problematic design features. Each housing was then 3D printed, assembled, and tested alongside a controlled setup for direct comparison. Each test ran for at least 24 hours, which allowed for random fluctuation in data collection for each sensing category. The results of each test were then compared to the control setup, and any previous noteworthy prototype. The testing and redesigning process was repeated multiple times to achieve the best possible performance of the device.
The latest design of the SpaceBot device housing produced a 103.7% and 112.9% improvement to relative temperature and humidity readings, respectively. This was achieved by compartmentalizing each sensor within the device while separating them from the heat source. Despite relocating the microphones relative to the particulate sensor, the noise mitigation improvements remain inconclusive. Adding aluminum heat sinks to the Raspberry Piâ€™s processors creates a significant heat flux away from them, which reduces their overall operating temperature by 13.6%.
Restructuring the housing, adding passive components, and reorienting the internal hardware drastically reduced the amount of error experienced during data collection, producing more accurate readings.