Using Environmental Sensors to Record Lives

This post from Mauro Cherubini made me start thinking again about a topic I have written extensively on, recording a life via technology. Numerous entities have made developments along the lines of enabling a person's life, particularly their digital interactions, recordable, searchable and linkable by capturing and relating e-mails, phone calls, text messages, digital images, calendars, etc. to piece together a record of that person's actions and interactions. Lately, a start-up called Dandelife got some press for a new application that helps link together dates, places, and images to create a digital social biography. Microsoft has also worked in this space, and doubtless Google is doing so as well.

What those systems don't capture are our offline contexts - where we went running this morning in Mauro's example, or what we were listening to when we were sitting in the park, or what the UV index was while we toasted in the sun on the beach. Maybe these aren't all items we want to capture right away, but over time the definition of context will expand to add as many datapoints as exist independently about our activities and locations as can be captured and linked. This of course means utilizing current disconnected sensors in a connected network, or finding a way through, say, a personal markup language or cross-referencing our personal IP address (one that moves with us as an individual) with location data to do a sweep of local sensors, grabbing environmental information on the fly. So some context 'bot would sweep networks of information looking for my personal identifier  and collect context from whatever sensors picked me to up relate to me. Now we are talking about the need for petabytes of storage. 

Would this be strange, dangerous or at least occasionally problematic? You bet. But that doesn't mean it won't happen. Personal mashups are on their way, and what's described above isn't so far off as context becomes a more all-encompassing term.