MAMAS: Mealtime Assistance Using Magnetometer and Speech Recognition

Children’s problemating eating behavior is one of the biggest problem parents suffer from. Picky eating is observed in 80% of children with developmental disorders and in 20-50% of those without. Many parents fill stressed out by their children’s mealtime behaviors.

Can we promote child’s heathy eating behaviors in an affordable way?

Preliminary Expert Interview & User Survey

We conducted interviews with three experts on child’s eating behaivors. Also, we conducted preliminary survey with 72 parents from local pediatrics. Survey questions included 1) general information, 2) challenges, 3) behavioral patterns, and 4) sentiments regarding mealtimes with their children.

MAMAS chart 1

Major challenges parents faced were as follows.

  1. They did not realize their own language habits.
  2. They had to distract children to feed them.
  3. They themselves were depressed and stressed out.

Surprisingly, we found out that critical factors of mealtime education was not about “children’s bad habits,” but about the followings.

  1. Positive parent-child mealtime interaction and
  2. Exclusion of all media distraction.


We decided to make an application called MAMAS, a mealtime assistant using magnetometer and speech recognition.

MAMAS backend diagram

Based on our user study, we designed the app to have following features.

  1. Track mealtime interaction patterns in an noninvasive way.
  2. Augment analysis with self-reported data and quantification.
  3. Provide data-assisted analysis for parents’ self-reflection.

MAMAS solution diagram


We are developing this MVP into a deployable iOS application for actual long-term deployment experiment.

p.s. The name MAMAS came from our favorite sandwhich place, Cafe Mamas.