eCUBE is an on desk assistant specially designed for the workers and students who have to force change their homes as their working environment due to the pandemic.
1. What motivates you to be part of Bounce Back Ideas?
The world is in danger. People are dying, businesses are falling down, unemployment is rocketing to the sky. I am experiencing this daily as I am from a developing country that is thriving to find its own way. It is the chance for the youth to step up and brainstorm new ways to lift the burden. So, I find Bounce back Ideas is a perfect opportunity for us to present our ideas to a global audience. When I saw this challenge on the challenge portal I got that instant feeling to join hands and do something innovative for the society.
2. Problem/Opportunity Statement:
Pandemic has changed the way of our lives. Everyone has to adapt their lifestyles and their work according to the constantly changing health guidelines. Some have lost their jobs. Some have to transform their work into a remote work-from-home environment. Some have chosen emerging online jobs. The most noticeable change the pandemic has done to the people is they have been forced to transform their own homes as their working environment creating a set of whole new problems. Now, they have to sit in front of a computer machine for more than 8 hours a day for their work. Some have managed to overcome this challenge but the majority of the workforce is not happy about this. They feel distressed and lonely. They are discouraged to do their work. Some have got physical injuries such as Upper limb disorders, eye strains, and other repetitive strain injuries as a result of sitting in front of a computer screen for a long time. Also, students find it hard to focus on their learning materials in this distant learning mechanism. Personally, being a student, I have faced a lot of problems in the online learning system as it is very hard to keep attention to the digital materials without any interactions with the physical world.
3. Short Idea Description:
My idea is to develop an on-desk assistant for the remote workers and also for the students. This is an electronic cube integrated with a number of sensors that can be plugged as standalone modules as per the requirement and a microcontroller. The name "eCUBE" is given simply because it is an electronic cube.. eCUBE will be a perfect partner and an assistant for the workers and students. eCUBE can detect the behaviors and the emotions of the user and will interact according to the mood of the user. Also, our little friend will help the user to adjust his bad postures promoting ergonomics in the online working community.
4. Long Idea Description:
eCUBE is a smart assistant specially designed for remote workers and students. It will help the users to focus on their work while being efficient and productive. eCUBE integrates concepts of embedded systems, machine learning, ergonomics and psychology to build a friendlier working environment for the user.
It has integrated sensors as well as add-on modules to detect user behaviors and emotions that are pre-trained with efficient machine learning algorithms. eCUBE is mainly integrated with an eye-friendly OLED display, a gesture sensor for easier navigation and a LED system for emotion representation. Also, it has a port to connect specially designed capacitive sensing plates which can be placed on the working table(preferably under the table mattress) and on the chair. These capacitive sensors will be used to detect the presence of the user, his postures, and his behaviors. As an example, this can detect if the user is sitting in front of the desk for a long time without taking a break and our little friend will simply alarm the user to take an interval. Recent studies have shown that taking short breaks while working will drastically improve performance. This capacitive sensor can detect whether the user is sleeping on the desk or he is leaning towards the computer monitor using a specially trained machine learning algorithm.
eCUBE is using a psychological approach to help users overcome work-related stress and anxiety. It uses color psychology, binaural audio techniques, and meditational music to uplift the user's emotional state. Also, eCUBE will automatically adjust the lighting and temperature levels of the working environment if the present state is not good for the user's mood and health. As an example, recent studies have shown different ambient light intensities, background music and temperature levels may have a positive or negative effect on one's productivity and healthiness. So, eCUBE will measure the current state of the user and his environment such as ambient light, temperature and user's behaviors.
eCUBE is built upon a low-cost Atmega 328 chip reducing the price of the product. It is using the microML package built on python and C++ to run machine learning models on this low power chip. So, the eCUBE will be an affordable and user-friendly device for the workers as well as for the students. Thw wifi module will enable IOT access for the eCUBE giving the full control over the smart home accessories likes Air Conditioners and lights.
5. Why is your idea innovative?
At present we can see a lot of personal assistants and smart home assistants have appeared in the market. (ex - Google Home, Amazon echo) They are mostly designed to cater general purposes and usually come up with a high price tag. They are using high-end services and will give users a better experience as an AI assistant. But eCUBE has a very different purpose and it is not an ordinary AI assistant. It is specially designed to support the remote workers as a mentor or a friend. Also, it will be very affordable compared to his big brothers as it not designed for smart home applications or AI assistance devices. Also turning a cheap Atmega 328 to handle robust machine learning algorithms is a huge challenge. So, this approach of using machine learning models on low power chips is innovative in its own way and will open up a new way of affordable smart products for the social benefit.
6. What are the potential impacts of your idea?
In this pandemic situation, we have seen a lot of companies are receiving financial losses. It is mainly because the online working is not giving the optimum performance out of the employees. Employees are working very lethargic demoralized way. So, it is very essential to create a good environment for the remote working people without letting them in danger during the pandemic. eCUBE will help to create a pleasant environment in their homes for the remote workers reducing their stress and loneliness. It will ensure their focus on work improving performance and efficiency. This will be really helpful to transform their working style into the work from the home method in a very comfortable way. Also, eCUBE promotes ergonomics reducing physical injuries due to wrong postures. So eCUBE will ensure both the mental and physical well-being of the employees eventually contributing to a performance boost in the company.
Also, eCUBE platform is very useful for distant learning students. It will remind them to take timely breaks based on the activities of the students making them to work productively. Also it will promote good sitting postures for the students as they are very likely prone to physical injuries due to bad postures. Meditative music and binaural beats along with intelligent changes of the light intensities and temperature will make sure the students will improve their productivity in learning more than ever.
7. How do you plan to implement this idea?
eCUBE needs physical equipment like Arduino pro mini, ESP8266 wifi module, TMG39931 color and gesture sensor, SD card module, speakers, capacitive sensors etc.
It should be programmed using C++ programming language. Machine learning algorithms will be trained using the Sci-kit library in python programming language. Trained models will be ported into the C language using the microML package hence reducing the RAM and flash requirements ti run robust Machine learning Algorithms such as decision trees and random vector machines. Training is the most difficult part of the development phase. We have to collect a lot of raw data from cap sensors and gesture sensors to output meaningful information using the ML models. So, we have to conduct researches with test rounds to collect data of different postures and gestures to feed the machine learning algorithms
8. Why is your team ideal to implement this idea?
I am learning electronics and programming on my own and I am very enthusiastic about developing new devices. At present, I am experimenting with deploying machine learning models on low-power memory-constrained embedded systems such as Arduino. As I am currently waiting to enter the premier technological university in Sri Lanka, I believe I will be able to complete this project using the expertise of different professionals in this field. Also, I have personally gone through the difficulties of remote working and learning. So I have a good idea about the reality of online working and distant learning. So, I am building eCUBE to overcome the difficulties that I have experienced before.
A 3d prototype of the eCUBE is made using the clara.io web UI.
Diagram of the eCUBE