As any parent I am constantly concerned about my child's health. With data it is easier to identify emerging problems and also diagnose the underlying problem as a precaution rather than merely reactive. For these reasons we spent the past year on our Quantified Baby project.
Throughout the year we have been maintaining various quantitative and qualitative data points of our son's habits and growth. From the day he was born we gathered every time he ate, slept, pooped, peed, and took a bath. This included the quantity of food, time of sleep, and even (optionally) the color, consistency and leakiness of his output.
Overall our son has been extremely healthy, happy, and effusive. To date his he only been sick with a common cold once for about 5 days, had no allergies, and slept through the night after the first three weeks when he was effectively gaining weight. While I know that our data capture and analysis doesn't account for our fortune, we do believe that it was beneficial to our own parenting, regularity, and ability to be aware and informed of his health for any doctor visits. We also didn't sweat all of the details. There are a few gaps due to lack of sleep, general distraction, or sometimes simply bucking the machine (i.e. "I don't wanna").
We learned a lot of insights along the way, particularly in the methods, benefits, and difficulties of quantifying your baby. For our measuring, we used the capable, full featured, and easy to use Total Baby. Designed by an engineer for his own family, it has the practicality of design by someone that is also forced to use it in a bleary eyed state with a crying baby in another hand, and a bottle/diaper/blanket in the other. This is not meant as a review of this particular application but more as a highlight to what the baseline and required features should be of any quantified self applications.
Make it Useful
Many 'quantified self' applications take a conceited view of their data capture and require you to enter data with little to no information value returned to you. These tools are fun at first but quickly become forgotten since they required action outside of the normal activity.
By contrast, quantified self tools must provide at least immediate value. By example our measurement of feeding provided a timer of the current activity, and also set a (configurable) alarm for the next feeding that would typically occur in 2 hours at the beginning. When you've been cycling through several days of polyphasic short sleeps, this type of simple arithmetic becomes harder than you may imagine.
So by providing an immediate benefit (automatic recurrence timer) we were clearly incentivized to keep using the application. A missed feeding would quickly get entered so we knew how he was doing throughout the days and a simple count of times and amount he fed to ensure positive weight gain. You can imagine the potential for existing social networks to measure and alert on emerging trends that may impact your health through continuous input for all of your data.
Make it Easy
While we are both engineers, when first learning to deal with another human life your focus tends to be on their needs and not learning complex applications. Actions and questions need to be extremely clear and operable with a single hand and thumb.
It should also be easy to get quick statistics from a glance so that you can take action if necessary. With our tracking we could immediate see the time since last feeding, changes, sleep, and even bath or other customizable timers. This bio-dashboard meant we were reassured through parallel mechanisms that we were tracking data and everything was on track with him.
Automated tracking is even better as they don't require remembering or only capturing observed actions. Deb Roy's work in 'The Birth of a Word' demonstrates passive monitoring of language development that would not be as clear in limited measurements. I had originally planned a few hardware sensors that would measure crib pressure for motion in sleep and rolling over - but as they say "finish all your projects before the kid is born; you won't get any done afterwards".
Make it Shareable
The data are useful to one person, but arguably are just a substitute for memory. However with two or more people, a quantified self application becomes the collaboration center-point to have multiple inputs and actions on the information. When my wife would change diapers or finish a feeding my own phone would update with these events and I would be continuously informed without having ask, or even worse, wake her up.
The data also worked on any of our iOS-based devices. So instead of having to chase down a single unique device we could instead grab anyone iPhone or iPad that we had nearest to us - which usually wasn't far as they are also great devices for reading while rocking, or playing some soothing music - proving the mobile phone in and of itself the greatest invention for parents.
Make it Open
A quantified self application should be expandable, and open to customization. While by default Total Baby has a good configuration of feedings, colors (if you need to be that specific), food types, and activity types - you are also able to add your own activities, timers, and as I mentioned before, automatic reminders. That means if you have special needs, medications, therapy, or just want to ensure validation that grandparents are getting equal time, then it's important that the application allow you to add these as necessary.
A surprising and welcome feature of Total Baby was that all of the data were exportable as a Spreadsheet CSV and SQLite database file. While not the cleanest data model (times are just strings in the event descriptions which are not enumerated), it allowed us to play with charting and general metrics as our son grew. The growth charts and other timelines are also all exportable and email-able to family, doctors or caretakers. I can also make backups in case any of our devices are lost.
Opportunities for Health
'It's your body, make sure you know what is going on with it.'
In personal health, I'm continually astonished by lack of personal data that is measured and retained. How many times do I need to fill out basic details such as address or blood type that I am dubious that real information such as blood levels, heart states, or other extremely important information is gathered, available, or used for any kind of passive analysis and alerting.
While I do trust my health practitioner, no one is as invested in the health of my body as I am. Doctors are busy and laden by antiquated technology that prevents them from providing the care they would really like to. There have been numerous attempts are revolutionizing the healthcare industry that fail through lack of focus, complex policy, privacy issues, or just lack of general initiative.
I want to see the same technology that reroutes traffic, recommends new shoe ads, and sorts my email to understand when I need to alter medications, change my behavior or visit a specialist who upon my arrival would have a full, and appropriate, analysis and access to my data history.
By focusing on the quantified self, we are empowered to monitor and maintain our own well being - and obviously with children to assist in theirs as they grow. I can invest the time, effort, and even technological development that creates a solution now that will aid my own family.