1. Mobisys Seminar 02.12.08
Sentient Computing for Innovation
Understanding Organisational Behaviour through Pervasive Technology
Irene Lopez de Vallejo, PhD Candidate
i.vallejo@ucl.ac.uk
The Bartlett, Faculty of the Built Environment
UCL University College London
2. Mobisys Seminar 02.12.08
Initial considerations
• Sentient computing is the proposition that applications can
be made more responsive and useful by observing and
reacting to the physical world Andy Hopper, The Royal Society Clifford Paterson Lecture,1999
– Built Environment
– Inhabitants
• Innovation key form of organisational knowledge creation,
organisations create and define problems and actively
develop new knowledge to solve them Nonaka 1994
– Individual and collective employees expertise
– Building design
3. Mobisys Seminar 02.12.08
Key question
• How can location tracking systems contribute to
the understanding of informal face-to-face
interaction in the workplace?
4. Mobisys Seminar 02.12.08
Interaction, Knowledge Management and
Innovation
• Physical interaction key issues for
today’s firms
• To create and transfer knowledge
individuals need to interact
• Informal face-to-face interactions
essential for
– Trust building
– Flow of ideas
– Social relationships original context on the page: www.ikiw.org/category/interdisciplinary/
5. Mobisys Seminar 02.12.08
The other side of the coin: Solo work
• Get work done!
• Concentration vs communication
– Fluctuating desired levels of privacy and interaction
– Individual creativity
• Interruptions
– Stress levels
– Time famine
• Paradox of knowledge management
– Best transfers serendipitous, personal and private
– Best insights need periods of intense reflection combined exposure
6. Mobisys Seminar 02.12.08
Organisational and Design considerations
• Work context intense, fast, novel and collaborative
• Create processes, structures and environments that allow
and encourage individuals to interact to promote informal
knowledge transfer
• Face-to-face interaction is a mechanism
• Design a strategic tool to support emergent organisational
forms
– Adequate time and space for interaction
– Open spaces are more favourable
7. Mobisys Seminar 02.12.08
Pervasive technologies to measure interaction
• Current methods
– Static picture
– Self assesment and observations
• A new tool
– Indoor location tracking technologies
– Dynamic and long term view
• A real life deployment:
– Ultra-wideband RF technology: Nationwide pilot
project
8. Mobisys Seminar 02.12.08
Pervasive technologies to measure interaction
If you cannot measure it, you cannot manage it
• Current methods
– Static picture
– Self assessments and observations
• A new tool
– Indoor location tracking technologies
– Dynamic and long term view
• A real life deployment:
– Ultra-wideband RF technology
– Nationwide pilot project
9. Mobisys Seminar 02.12.08
A New Method to study interaction:
background
• L-T system calculate position
accuracy 15 cm, 48% time
• Spatial and social challenges
• Raw location dataset
– Noise
– Assumptions
– Definitions
– Mathematical Manipulation
10. Mobisys Seminar 02.12.08
A New Method to study interaction:
background
Can interaction be measured, understood and managed?
• L-T system calculate position
accuracy 15 cm, 48% time
• Spatial and social challenges
• Raw location dataset
– Noise
– Assumptions
– Definitions
– Mathematical Manipulation
11. Mobisys Seminar 02.12.08
Dealing with noise
• No noise free location technology
• Kalman filter
• Delete all XYZ = 0
• Delete entries Z < 0.5, > 2 metres
• Delete entries negative X and Y data points
12. Mobisys Seminar 02.12.08
Preliminary assumptions
• Set of events that exclude each other
• Each event has two states: ON and OFF
• Time line
– Organisation
• Spatial setting
– Deployment area
15. Mobisys Seminar 02.12.08
Definitions: Spatial factors
• Cartesian coordinates
• Pre-allocarted zones
• Configuration of office space
16. Mobisys Seminar 02.12.08
Summary defintions and unit of analysis
Behaviour
Measures SOLO INTERACTION OTHER
PERSONAL 1 individual 2,3,4,5 and 6 individuals Outside
SPACE Aggregation of workforce Aggregation of workforce deployment
No trespassing personal Overlapping of personal Aggregation
bubble of 1.5 feet radius or bubbles for a minimum of workforce
overlapping of personal duration threshold value
bubbles of under 30 seconds of 30 seconds
TEMPORAL Occurrence, recurrence, duration, period or cycle, interval and rhythm.
FACTORS
LOCATION Cartesian coordinates (XYZ)
FACTORS Pre-allocated zones1
Configuration of office space
22. Mobisys Seminar 02.12.08
Q& perhaps some A
• Dynamic visualization of human behaviour in organisations
• Predictive model
• Location systems not the drivers
• Design and organisation studies