DAT602 – Project Idea – uSense

The Project Brief

In this assignment, you are asked in groups to develop a physical “intelligent” object for the Future Home.

You have the freedom to select any tools or resources in accomplishing your goal, such as: sensors /electronics / microcontrollers (Arduino, NodeMCU, Raspberry Pi, LattePanda), coding platforms (HTML, CSS, React, Node.js, JavaScript, Node-Red), databases (MongoDB, SQL, Firebase), PaaS/IaaS/SaaS (Amazon AWS, Heroku, Azure, IBM), visualization and computational media (Unity/Unreal, WebGL/Three.js, D3).

However, you are constrained by the following:

1. The object/system that you create needs to be part of the context – home / domestic life / personal space.
2. The object/system has to add something significantly new and innovative to the main context rather than (a) provide a device that uses a sensor to trigger a specific elementary outcome, or (b) replicate an already existing object/system.
3. The object/system has to be fully functional and demonstrate exactly what it is intended to do according to your main idea/concept that you will propose.
4. You need to be methodological, work efficiently as a group, and demonstrate professional practices – documentation, schematics, working progress. It is necessary to use blogs and GitHub to show your progress throughout the module’s duration.

Our Project Idea

On 10 October 2018, I created a Facebook group for use by  myself, Lee Lashbrook, and Sam Williams. This group could be used to share ideas, thoughts and progress on an informal basis.

The following day, via the Facebook group, I proposed the following project idea:

“I believe we might be doing some brainstorming in today’s session but I thought I would share a possible idea … it’s a ‘Mood Board’ that detects the sentiment in a room/house via voice recognition (maybe!) and adjusts lighting/music/images appropriately. So if people were feeling stressed, the lighting might change to low level/the music might be soothing or sounds of nature/the images might be calming, etc. We could develop and small prototype – a single screen, some LEDs or Neopixels and a speaker. Just an idea.”

After some further thought I added:

“We could possibly bring in some issues around mental health and well-being, which is pretty current.”

And finally:

“Also just been looking into emotion detection using facial recognition. I’ve previously done some work with the Kinect for gesture recognition so this might be an improvement/addition to the voice recognition.”

This proposal was accepted by the other members of the group and, during a group work workshop at university, we had some initial discussion around the idea, functionality and implementation.

During this workshop, our tutor provided some valuable feedback regarding research and some technologies that we could perhaps utilise, such as IBM Tone Analysis Service and IBM Watson. Our tutor also suggested that we consider the idea of using an individual’s social media posts to help determine their mood through sentiment analysis.