Indoor Space Utilization
SOM Research & development 2017
A Passive System for Quantifying Indoor Space Utilization, in the proceedings of the ACADIA conference Disciplines & Disruptions (ACADIA 2017).
I initiated a self-directed research project at SOM to create a prototype of a new sensing device to automatically and anonymously evaluate space utilization in the workplace. After writing custom tools for the architectural design process, I realized that none of them addressed the end-user experience. I viewed this as an opportunity to develop new tools to inform architects about the social implications of their designs. I developed a prototype to expose new aspects of the occupant experience through data capturing movement, room occupancy and temperature.
I was independently responsible for all aspects of the project including ideation, research, development of the electronic hardware, programming, and data analysis. I defined my design objectives as follows:
Small Form Factor: The device must be small and lightweight for simple, unobtrusive installation in any space.
Scalability: The device (or devices) must be scalable to rooms of different shapes and sizes.
Communications: The device must be capable of wireless data transfer.
Power: The device must have low power consumption and a long battery life.
Occupant Anonymity: The device must maintain occupant anonymity. This is NOT a surveillance system.
I considered a variety of sensors that could differentiate between humans and their non-human surroundings without jeopardizing the human's identity. After considering infrared cameras, range finders and break sensors, I chose to use an infrared array sensor to collect extremely low resolution thermal images. Compared to the competitors, this sensor covered the largest area with the highest density of data points for the lowest price, all while abstracting humans to unidentifiable clusters of pixels. I designed a custom printed circuit board for the sensor and connected it with a WiFi-enabled microcontroller.
I wrote a blob detection algorithm in Python to filter the data and isolate the human figures within the viewing area. I tested my system on two scenarios: counting stationary people and tracking a single moving occupant. I used the processed data to generate simple diagrams of the results.
Although a very early prototype, this system successfully captured human occupancy and movement within indoor spaces. In the future, the device will need further refinement for battery consumption, high density room occupancy, and better localization accuracy. I envision that this device will be used as a tool for architects to receive feedback about how their designs influence social interaction, circulation, productivity, and user comfort for indoor environments.