Your Personal Emotion Tracker & Well-Being Guide
Stay in tune with your emotions and improve your well-being with our smart emotion tracker! It helps you:
-
Track your mood over time to understand patterns
-
Get personalized activity recommendations to boost happiness
-
Stay balanced with guided activities like music, reading, travel, and exercise
Whether you're feeling low or just want to maintain a healthy lifestyle, this tool supports your emotional wellness every step of the way!
About the Project
Emotion Tracker was started as a Fun project in Hackathon while working with Siemens around Oct-Nov'22. As we were working on a family of applications the idea was to bring Emotion tracker as a fun tool to all our customers. Ideas got well appreciated by the Jury. We even applied for patent rights for this tool in 2023 ( Patent ID : 2023E02602 ).
Background
Idea evolved during the time of Hackathon, after many appreciations and a good lot visibility in the organization. We have presented the same to all the stakeholders from other verticals. Since Emotion tracker is a generic tool and can be integrate with any product or vertical having users from across the world. After multiple demoe's to Global stakeholders
it was approved to go with all the products in the future and we were in a game to enhance it to produce more insights
with the help of the data that we were receiving from the user.
Requirements and Expectations
Requirements were very simple, Emotion tracker tool can be integrated to any Siemens tools, user can enable from a settings page of the respective application and if any user wants to enable the tool they need to acknowledge the privacy policy for using their personal data to bring insights. Currently, tool will capture the user data only when the user clicks on the floating icon on the app and navigates to the landing page of emotion tracker. Only from then the app will start capturing data.
​
App is intelligent enough to process the data with the NLP's and provide recommendations as what can be done to stay Happy.
Functional Flow of the Proposed model
System Architecture
System Architecture According to the user's emotions, the front-end receives various forms of inputs. In order to classify the user's emotions accurately, the model integrates three different techniques: image selection, sentiment analysis, and category selection.
The steps involved in emotion recognition and solution recommendation system
Resolving Stakeholder Conflict: Building Trust with a Emotion Wheel
There was a serious conflict between Sub segment head, Project managers, Architects and Me. Stakeholders think that displaying a wheel have no significant use for a user as it just blocks the camera view.
​
My point is when the wheel is displayed with the actual pointer and the key words that we fetch from the wheel and showing on either side of the cam area will bring trust and faith in the data that the tool is providing. Users can believe that the tool is operating with clear set of data points and brings authenticity in the information. Eventually users will also start considering the Tool's Recommendations seriously.
​
If there is no wheel but just the plain camera view with those widgets in the bottom to track several emotions may not bring the authenticity as it may also perceived as a fun tool and not to be taken seriously.
​
After multiple discussions, stakeholders got convinced with my reasoning. So we considered both the scenarios so came up with the idea of having two modes. One with Affect which is by default selected when the wheel is displayed. Second with the Affect is not selected then the wheel will be disappeared from the screen.
This is how the app looks by default
App provides two modes, one with Affect and one without Affect.
By default Affect is selected and the emotion wheel is visible on users camera along with the key words that are fetched by the models and displays to the user.
​
Second mode is without Affect, when unselected Affect the Emotion wheel will be disappearing from the screen along with the keywords. It's a plain ground where the user can only see themselves on camera and track the stats and recommendations.
This is how the app looks without 'Affect'
Prototype
Link for Prototype
You can refer the prototype here - Take me to the Prototype
Next Steps
Leveraging the same data, the plan was to scale the idea when a user uses our applications and track the emotions based on the use case. Of Course to achieve this we needed a much broader audience, stakeholders, Project Heads because to make it real the applications will have to track the user emotions all the time which makes the majority of the users feel the tool is hampering privacy and other compliance issues etc. Many challenges were seen in that front but we have kept our proposal and if we get a go ahead with the list of challenges from all aspects then we can start thinking in depth to address each. Meanwhile we moved to a different project.