Instrumented living environments




Data sets generated from instrumented living environments



Source
Scope
Sensors
Researcher's
Home?
Annotation
Availability
Intel Berkeley
4 houses, 3 weeks each
UbiSense tags
No
Sparse self-report data
Unknown
Great Northern Haven, Dundalk, Ireland
16 purpose-built smart apartments occupied on a full-time basis by older adults since June 2010 (real-world continuous data).
Over 2,300 sensors in the development including motion sensors, contact sensors, brightness sensors, temperature sensors, electricity sensors, water sensors.
No
None
Open to research collaboration
MARC
1 house, 12 weeks
Motion/switches
No
Self report from PDA
Unknown
Intel Seattle
Subjects perform activities on list
RFID/bracelet
Yes
Researcher observation
Unknown
Intel Seattle, U. Washington
One author performs activities on list, 10 days
RFID/2 gloves
Yes
Researcher observation
Unknown
Intel Seattle, U. Washington
3 subjects perform activities on list. 1 hour of data
RFID/bracelet
Yes
Researcher observation
Unknown
Intel Seattle, U. Washington
2 subjects perform activities for 12 hours
Acceleration, light, temperature, barometric pressure
No
Researcher observation
Unknown
UC Berkeley
13 male and 7 female between ages 19 and 75 perform repetitive actions in 13 action categories.
Accelerometer and gyroscope. (EIP, ECG sensors planned)
Lab
Category annotated
Yes
Hiroshima U.
1 house, 6 months
IR occupancy
Unknown
None
Unknown
Waseda U.
3 houses for one month each (3 subjects)
Switches, IR occupancy, temperature, humidity, carbon dioxide
No
None
Unknown
CMU
1 house, 5 days
24 motion detectors, 24 door switches
Yes
Self report room location
Unknown
CMU
1 home, 6 weeks
Acoustical on pipes
Yes
None
Case by case
GaTech
1 year
16 floor pressure
No
6 weeks subject diaries
Unknown
Curtin U. of Technology
One author performs 30 examples of single activity (Breakfast)
Switches, pressure mats
Yes
Researcher observation
Unknown
MIT
Weeks
MITes: Binary switches
No
Yes
Yes
MIT
Test datasets available; 2.5 months to be released
PlaceLab: Multi-modal, 800+
No
Yes
Yes
U. Rochester LACE
ADL simulations: 5 participants, 2 sessions, 6 activities, 5 trials
Video and RFID
Simulated home
Unknown
With Approval
Washington State U. CASAS
Varies
Motion sensors, door sensors, VOIP usage, object sensors
No
Researcher observation
Yes
U. Tokyo ISIS
Unknown
Motion capture
No
Researcher observation
Yes
U. Amsterdam
1 house, 4 weeks
Wireless sensor network of reed switches and water level sensor
Yes
Self-report through bluetooth headset
Yes
CMU QoLT
30 subjects cooking 5 recipes (brownie, sandwich, eggs, salad and pizza) in a kitchen
Video, Audio, Motion Capture, IMUs, bluetooth accelerometers, RFID/iBracelet, eWatch, Bodymedia
No
Yes
Yes
UF / Washington State
Synthesized data set version of Washington State CASAS (with comparison)
Same as CASAS
No
Researcher observation
Yes
UF
No dataset. A specification document of a proposed standard to represent datasets. Initiated by UF but we call for participation by workshop participants.
Document title: SDDL: Sensory Dataset Description Language.
No
-
Yes
U. Edinburgh
Video dataset of detected people walking through the atrium of a building, where the data comes from an overhead camera.
Bounding boxes and colour histograms for moving targets; lists of detections and splines
Public space
No
Yes
LIG/Multicom
1 flat, 2 sets :
  • A set on comfort evaluation, with 20 samples including users inputs
  • A set with one year of various data
Presence, user interaction, indoor metrics, energy measurements, etc
Yes
Users self perceptions
Yes
TIMC-IMAG and LIG
1 house, 15 participants perform activities of daily living
Microphones, presence infra-red sensors, webcams, thermometer and hygrometer, home-made worn accelerometer and magnetometer
No
Researcher observation
Yes (more with approval)