Instrumented People, Places and Things: device ecosystems, 2011 to 2012
Start here
Living with devices. For a technology client's hardware group, the Iota team studied how people live with their laptops, phones and tablets, as what it called a personal technology ecosystem: which devices and accessories are called into action for a goal, in a context. The anchor is things. A pilot in the summer of 2011 came first; the fieldwork, from November 2011 to January 2012, instrumented 18 people for about 15 to 21 days each.
A proof of concept. In TJ's words: “The study is more proof of concept and the viability of the approach. More qualitative than quantitative.” The question was whether instruments on people's own things, logs on their software and the team's ethnography could together show how a person's devices live together. What follows is that picture, in cases and patterns from 18 people. It is not a measure of how people in general use their devices.
The team, in August 2011: “Our overall hypothesis is that there are three ways to approach an understanding of ecosystems, each of them different in what aspect we choose to ‘fix’ while allowing the others to vary. We can anchor the data in rich and dynamic descriptions of People, of Things, or of Places.”
The model of activity. The team crossed continuity with activity. Staying on one activity or related ones, continuously, is attending. Moving among unrelated ones, continuously, is tacking. Use that is not continuous is darting. Continuous meant filling the application logger's five-minute bucket.
The team judged “related” by eye. So that the model can run over every bucket, this page adds a rule: a bucket is continuous at 295 s, and one activity when one category holds 80% of its time. Its commonest mode matches the team's own label on 11 of 15 windows. The Model view moves both thresholds.
Moments
A draft, November 2026. Participants appear under codes assigned here, by study day, with no dates, names, titles, sites or places.
How it was drawn
- Days
- One row for each day of each participant, midnight to midnight. A blank row is a study day with nothing logged.
- A day
- Drawn as the team's labeler drew it: a lane for each logger; purple means moving, deeper for stronger movement; the red band means the logger held valid data; hour lines, heavier at noon. The application lanes above it are added: seconds in each five-minute bucket, by category, with the bucket's mode beneath.
- A use
- A row in the application logger's report: one application or document, with its seconds in one five-minute bucket. A 40-minute video is eight rows of 300 seconds, so a use is not a session.
- Things
- Drawn as the team's map of connections drew it, from the configurations it coded from the diary. The team's tool is the reference; point at a thing to light its links.
The instruments
- Loggers
- TJ's accelerometer loggers: an accelerometer, a clock and a memory card on a small board, logging about eight times a second after a vibration and about every 10 seconds otherwise. 67 of them, on laptops, phones, tablets, keyboards, mice, a headset and bags.
- Application logger
- RescueTime on laptops, Android phones and Android tablets. iPads and iPhones had the accelerometer only.
- GPS
- A receiver on each person. The team coded its tracks as home, work or school, other place and in transit; only those codes appear here.
- Bluetooth
- Fourteen loggers, on people and in rooms, and six beacons on things.
- Device logger
- A program on five laptops: what was plugged in, and mains or battery.
- Self-report
- Interviews, photographs, a mobile diary and text-message surveys, by the team's ethnographers. None of their text or pictures is here.
- Training sets
- In January 2012 the team recorded the logger on a tablet, a mouse, a laptop and a phone, doing known things, to see what each looks like to the accelerometer. The Training view replays them.
What it can and cannot tell
Eighteen people, recruited by how they used tablets and accessories, over two to three weeks: a model built from observation, not a measure of a population. “Fills the bucket” can only be read at the logger's five minutes; nothing shorter is seen. The rule's categories are the logger's, so “related” here is a category, not a judgment. Movement says a thing moved, not that it was used, and a laptop on a desk moves little. Two people have no application logs, and two of the four given new tablets have none on the tablet. Three people's loggers hold no labeled movement for their study days, and four people have no place codes for theirs. The Totals view gives what the logs add up to, each figure with how it is read.
The diary's text and the survey answers are not among the files read here. The Things view uses only the configurations the team coded from the diary.
The study was the Iota team's, for a technology client. TJ, Iota's CTO and at the time its only technologist, chose the hardware and software, designed and built the logger, wrote its firmware and built the tools that brought the data together. In TJ's words: “I did not do the interviews or the more traditional ethnographic research, I did collaborate with them as we invented new practices.”