Today's minimal virtual spaces make little room for a person to vent their emotional state on their virtual environment. Moodring 2.0 changes this by allowing the wearer to influence their virtual space according to their mood color. Moodring 2.0 is a wearable device which senses hand temperature and determines a mood color based off research on the emotional content of color. The wearer can then change their virtual space according to their mood color. For example, suppose the Moodring 2.0 wearer is upset at their boss, and sits down to start typing a notice of resignation expressing all of their pent-up aggression. Moodring 2.0 alerts the wearer to how their mood is influencing the tone of the letter by changing the font color to red meaning the author is upset. This font change is a visual cue to the wearer which allows for the wearer to reflect on their emotion and consider how they intend to employ their emotions in their daily activities.

Challenges & Tradeoffs

Using color to cue the wearer that they are in a decisive moment can lead support powerful decision-making tools as demonstrated by Dhairya Dand's Alcohol-aware Ice Cubes. In order to create a device which could use emotion awareness to impact decision-making we needed to figure out a way to imbed our technology in everyday activities. Affective computing technologies which claim to sense the wearer's mood make difficult tradeoffs in this area in order to detect mood. For example, in the early 90's the affective computing group under Rosalind Pickard experimented with a similar hand-worn device, called the galvactivator, which informed the wearer of changes in their physiological response through an LED display. The group used galvanic skin response (GSR), a measure of skin's electrical conductivity with links to emotional states, in order to determine what the LED would display. However, there are only a few places on the body where GSR can be reliably read, and the makers of galvactivator opted to place the sensor on the palm of the hand for a better reading. This tradeoff made the device glove-like inhibiting the hand's sensitivity to the environment and potentially impeding other tasks like typing. Moodring 2.0 avoids this unideal placement by opting for sensor placement at the base of the finger on the inside of the hand. Moodring 2.0 allows for the wearer to barely notice their mood is being sensed by incorporating all of the technology into the size of a ring.

Because of the ring's small size, making the device wireless was a significant challenge. We were pushed to include the Arduino because of the need to convert the temperature to color, but more importantly, to ensure the computer received the mood color from the ring. We also found that our emotion color model encountered significant interpretive challenges. While the model we relied on simplified emotion and color to precise equations, its multidimensional nature did not translate neatly to the single temperature scale. We operationalized our design based off data on arousal since this value corresponds most closely to GSR to arousal mappings. In order to incorporate the other emotional dimensions, we chose our color scale based off other dimensions and simplified the scale to three colors. One of the persistent challenges Moodring 2.0 continues to face is its sensitivity to environmental temperature. While Moodring 2.0 successfully registers a change in the user's temperature verses that of the environment, the fixed temperature to emotion scale does not respond to the impact of environmental temperature on the wearer's baseline temperature. For example, while Moodring 2.0 registers the wearer's temperature as five degrees above the environmental temperature it cannot differentiate between change due to the wearer being inside/outside or calm/upset.

Finding Mood Color

Similar to the galvactivator, Moodring 2.0 maps sensed values of the hand to color values based on research into sensed arousal. Inspired by the SPRWeb Recoloring tool, Moodring 2.0 examined the work of Patricia Valdez and Albert Mehrabian to determine appropriate temperature to color mappings. Valdez and Mehrabian researched the effects of color on emotion, Moodring 2.0 determines what colors are best representative of physiological response. We acknowledges color is a dubious representation of mood, but we still wanted to explore what you could do if mood rings truly reflected emotions. Using Valdez and Mehrabian's study of hue's effect on arousal, we determined which colors were indicative of peak, nadir, and mid-arousal values.

Figure from Valdez and Mehrabian's study on the effects of color on emotional response.

Using arousal values we created a similar scale for temperature with Purple-Blue being the lowest temperature value, Red the mid-value, and Green the highest. We then looked at all of Valdez and Mehrabian's data for the affects of color on pleasure, arousal, and dominance values in order to figure out which colors were representative of which emotions. We then derived a pleasure, arousal, dominance model for each color. Taking the median values of the data, we assigned positive values for each color's dimension if it's score in that dimension was above the median and negative for those lying below.

Valdez and Mehrabian's data. We normalized the data, and then we created a threshold based on the median. A horizontal line indicates where we drew the threshold.

We then determined the emotion for each color based off of Mehrabian and Russell's PAD emotion model.



Moodring 2.0 is made of:
  • 3 330ohm resistors
  • an RGB LED
  • a temperature sensor
  • a pipe cleaner
  • foam board
  • thread, ribbon
  • Velcro
  • a number of connecting wires
  • an Arduino Uno

Right now, Moodring 2.0 is connected to an Arduino board attached at the wrist. We hope to make Moodring 2.0 wire-free in future iterations. Once the ring was mocked up on the Arduino and breadboard, we came off the breadboard by soldering the materials together and condensing the circuit to the size of a ring. We then enclosed the wiring in a foam casing, and stitched together the encasing.



GitHub Code


1. K-Tell Commercial
2. Music used in the Video: