Group member: Dian Chen, Chad Underhill, Elsa Luthi, Meixintong Zhao
Introduction
As technology continues to advance, technology designs not only help people finish their works, missions efficiently, but also they have the great promotion for the pursuit of happiness. Wealth has been the most commonly used proxy for psychological wellbeing, which people like economists often argue that the link between wealth and wellbeing is weaker than we’d like to think. Therefore, basically the idea of this design is that using technology to not only get something done, but also promote health and wellbeing. In this design, we are going to use Affective library which is the program detector being able to recognize and analyze the face emotion.
Our motivation is going to help students in their daily studying life by analysis their face, emotion. Imagine someone is working at the midnight, he is really tired but still have to finish the assignment. We want to design the program to help him fight with the sleepy. This design can be running at the backstage on the computer, and the camera will detect the users emotion and the change of their face. When the detector realizes the user is tired and falls asleep, it will make the noise and pop the notification to cheer them up.
BrainStorm
After we decide what we are going to create for this design, we make the decision what we want our design to do. First, we have some ideations on this design, which it should be able to detect when user is falling asleep while studying and wake them up; detect when user is falling asleep while studying and tell them to take a break/nap, which also should have different response every time. In order to get attention from users, we decide to make some noise while asking them whether they are truly tired. Then, we want to the response is a confirmation tab which enables users to interact with computer. Therefore, if they do fall asleep, the program will give them some advance, and click on the tab is also required to pause the noise.
Final Prototype
This design require us to use JavaScript to create the webpage, which is in order to apply the emotion detector. However, our first mission to find a way to make our detector to recognize the “sleeping face”. The Affective library provides a sample of the eye closure, which is just able to detect when you close your eyes. Initially, this detector is quite sensitive, because no matter you are sleeping or blinking, the detector will always find you are closing your eyes. To deal with this problem, Elsa and Meixintong make some change on the original code for eye closure detector, which we try to make the detector have the reaction only when the eyes close for constantly 5 seconds. When the users eyes close for 5 seconds, the detector will determine the user is sleeping, which will trigger the following program. Once the user is sleeping, the program will make some noise to get their attention while a tab will be popped out to ask the user whether they are sleeping. If they give back the response, the noise will stop and some tab pops out with some advice. I set up several different advice for user, and make them randomly show up with the confirmation tab. Our team member Chad who is really good at website design makes our webpage relatively beautiful, which there are beautiful pictures with some dictums showing at the main page randomly. Also, he makes our the appearance of our advice really pretty and clear enough to get users’ attentions.
On the demo day, there are several users testing our design and provide us some useful advice to help us to continue to improve it. The most positive feedback is that almost everyone thinks our design is really useful for them and it works well, which they all need it in the midnight while studying. On tester wonders that how much CPU does our design take, because since the users need to open it for the whole time while they are studying. After we check the system information, it will take nearly 10 % which I can’t say it a small number because maybe the user will do something huge. We are not sure whether it will have a big influence on their work efficiency. Another user thinks that we should set the volume of the noise really loud, because for now, it is just a normal sound and hard to get users attentions. Also, we didn’t choice an aggressive sound to be our noise, which we thought that may suddenly shock our users. They also like our idea on randomly give the advice, which makes this program be more ‘careful’ to users. One of the big problem is that because the detector can only detect users’ eyes, when they are really falling asleep, their eyes will not be directly looking at the camera which the detector will fail to test it. An interesting thing is that Chad chooses a picture of water on the main page, and our professor thinks it makes him more want to fall asleep. There is also a mistake on our design on the demo day, which because we set if the user already falls asleep more than twice, the advice will only be ‘You should go to sleep now’ without playing sound. However, our sound will not play anymore after each two tests, which we already fix it. We also consider adding some tips on how to better users studying experience.
Summary for testing
Advice:
1. Can’t work pretend to fall asleep and what happens
2. Should store the data for future analysis
3. Care about how much CPU does that take
4. The volume should be loud, and an aggressive song choice
5. Sound should play on infinite loop(already fix it)
6. Some onscreen feature that was really dynamic
Improvement:
1. Add tips to how to better your studying experience
2. Able to analysis users’ health condition
3. Make the detector more flexible
Conclusion
For this design for well-being, I believe we did successfully design which helps users to keep their mind clear in midnight studying. Results after the final prototype testing provide us plenty of advice and a clear direction on what we are going to improvement. If there is another chance to make this design again, I will add more function on this sleepy alarm which will record data for each time users use it, and analysis their health condition. We want to create a deep relation between technology and human well-being, which should start with make the technology cares more about human emotion.
This is a demo video of our design: https://www.youtube.com/watch?v=c6WkLkdIX5Y&feature=youtu.be&app=desktop