Solution 1:

Build a wearable device that translates the child with Down Syndrome’s thoughts into a machine that says what they are thinking.This could increase the communication between a child with Down syndrome and their immediate environment. How the device is going to translate the child’s thoughts is unknown to us so far, but we believe that including one’s thoughts can improve communication, providing the child a base for what communication looks like. The main result of this solution would be that because the child is “speaking” more, more conversation  would ensue, thus resulting in increased speech input. Hopefully, the child would outgrow the device eventually, as it could be problematic if the child learns to rely too heavily on the device.

The following feedback was most helpful in refining solution #1:

  1. “Yes, and consider at what point to have the child not be on the device anymore”

We are definitely thinking about this as we progress through our moonshot solution and we think this is a very important point. We do not want to focus on a specific age as every child’s language development will vary. However, we think that when the child starts to demonstrate that the device is interfering with their ability to communicate (i.e. they begin speaking over the device) the device could indicate to the child’s caretaker that it is time to shorten the frequency of the use of the device.

2. “Yes and this may result in developing better grammar and a wider range of vocabulary.”

This is a wonderful point, exactly what we are striving for! Eventually, our hope is that the child will develop better grammar and vocabulary skills to the point where he/she no longer needs the device as much as before. That way, he/she does not become too reliant on the device while still improving the necessary skills to succeed.

3. “Yes, and you could also have the machine translate what they are thinking into text, in case they're in a setting where it's difficult to hear.”

This is a good option because this would also allow the child to practice language without needing to speak. Having this thoughts-to-text option would provide the wearer with a secondary mode of communicating as well.

4. “Yes but this would be a big hassle to use, carry around, and keep charged.”

We agree with this for sure. We were hoping that the device would be small enough to be portable but also large enough to show the translation of speech input into a text. It would resemble a fanny pack or cross-shoulder bag so that it doesn’t need to be carried around. The device would be designed so that it automatically connects to and is charged to a portable power battery that is small and easily carried around.

5. “Yes but will the child be able to filter out what is being said at all”

This is a really good point. Because the device would ideally speak “for the child,” we are implying that the device would act inline with the desires of the child in terms of what speech they want to produce. That is, the child would be able to filter out what is being said because the device would be “aware” of what the child wants to say and doesn’t want to say. Of course, this ability would be a feat of technology and is somewhat far-fetched (as moonshots should be!)

6. “Yes but... could this have a negative impact on their own ability to communicate with the outside world if the machine is speaking for them?” / “Yes but how will you prevent dependence on the device”

There were multiple “yes, but” comments on this exact point. It is hard to tell which individual child will be more or less dependent on the device over time, and it has proven to be a large barrier in our solution process. We are still not sure of how to respond to this question as dependence on the device is very likely to happen especially over longer periods of time.

Solution 2:

Build an AI that tracks the child’s language development and their speech intake via microphones and informs parents of the most effective ways to speak to their child in terms of increasing language skills.We think this is a great solution because it allows for increased communication between the child and the parents. This seems to be the most authentic way of increasing speech input, as an assistive wearable device is not required. This would also be a great option because it would increase speech input and facilitate conversation practice between child and parent, even at home.

The following feedback was most helpful in refining solution #2:

  1. “Yes and you could also build the database for the recommendations based off of researched techniques. For example using a lot of words that help the child with phonemes.”

We think this is a really convincing point. If we are able to build a database with all of this information from many different users, the power of the AI would be exponentially increased, thus yielding better recommendations for caregivers.

2. “Yes and this allows for an easy noninvasive approach.”

We think this is a really good point. Noninvasive methods are almost always a safer option because they pose less risk, which is why we feel more strongly about this solution.

3. “Yes and since AI will be used I think it will be able to dynamically change the type of conversations the child needs to have through changes in their language development.”

We think this is a great point. The AI would theoretically be able to adjust the suggestions it makes to parents as the child continues to grow, thus constantly providing the most appropriate advice in terms of how to engage in conversation with the child at any given age/developmental period.

4. “Yes but not everyone has parents”

This is the only “yes, but” that we received for solution #2. We think this is a great point. We will modify our language to refer to caregivers instead of parents in order to be more inclusive. This will include parents or whoever is the primary caregiver and guardian.

After reviewing the feedback we received, we have decided to focus and move forward with Solution 2. From the “yes, and” feedback we realize that creating some sort of database or word bank that could suggest improvements in different types of conversations would be really valuable and we think this would not be difficult to incorporate into our solution. From the “yes, but” feedback, moving forward we will not only continue to include more inclusive language, but think about how to make the AI solution most accessible to a wide range of children with Down Syndrome.

We rejected Solution 1, because we had the following concerns based off of the “yes, but” suggestions: an unhealthy dependence on the device and ways to prevent this; inconvenience with wearing the device plus its portable battery; and how the device would stand out, drawing unnecessary attention to the child, further harming their overall well-being.

Minor revisions to this solution would include: collecting information to build a database with research updates and effective techniques; a noninvasive approach only; an adaptive AI that adjusts to suggestions throughout the child’s development trajectory; and modified language catering to caregivers in case parental care is absent.