Inspiration

Our inspiration for this project came from a shared passion for sustainability and efficiency within the workplace. We wanted to create a website that would not only monitor, but also alert people to their bad habits that are commonplace in every home, office and public building.

What it does

Our project is named 'Sense & Sensibility' as it serves a dual purpose. The initial 'Sense' stage ensures that several monitors (Temperature, Light, CO2 and Motion) are applied on a live floor-plan accessible on our website. The floorpan pulls data sets from these sensors and allocates an overall score out of 100. Score thresholds are assigned colours and each room is monitored on an individual scale to assess its efficiency. A key feature that separates our project from other simple monitoring websites is our unique messaging function. Each office space is assigned to an individual and that person is responsible for their own room. Bad habits, such as leaving the window open whilst the radiator is turned on, will be logged on our website and the user will receive a friendly reminder via SMS message. This reminder will prompt them to think more sustainable and suggest alternative ways to cool the room down in this example. Our goal is that office spaces can use this website to create incentivised competitions that not only reduce building running costs and increase efficiency, but do so in a way that is fun, easy to access and simple to use.

How we built it

We constructed a client-server system to process and deliver relevant data to our users. The backend server consisted of a data processing pipeline to handle information received from various sensors located around the building. We constructed a dataset of dummy data using a generative deep learning auto-encoder trained on real data to learn the distribution of the sensor readings to produce an accurate sample. We constructed an algorithm to assign scores to rooms based on the various sensor values to ascertain the level of sustainability. The front-end client was constructed using ReactJS, MUI, and P5 to provide a seamless and informative interface for the user.

Challenges we ran into

Scaling graphics. Network interfacing. CIS. Bounding Box.

Accomplishments that we're proud of

Bounding Box collision detection. Constructing a functioning prototype in the limited time available.

What we learned

How to interface between numerous APIs and combine them to produce a more comprehensive project.

What's next for Sense & Sensibility

Bed

Share this project:

Updates