tapping into the wells of social energy: a case study based on falls identification
TRANSCRIPT
Tapping Into the Wells of
Social Energy:A Case Study Based on Falls Identification
Vincenzo De Florio and Arianit PajazitiMOSAIC/Universiteit Antwerpen & MOSAIC/iMinds
UNIVERSITY OF ANTWERP
MOSAIC Research Group
"We are greatly frustrated
by all our local, static organization
of an obsolete yesterday."
Richard Buckminster Fuller,
Synergetics I
It is the axioms that make up the system…
• Regardless of its nature, any system is affected by its
design assumptions.
Our organizations and services are no exception.
• “Wrong” assumptions inefficiency & fragility
Challenge
• Being able to (safely!) mutate our organizational paradigms and assumptions
• Move from
“a local, static organization of an obsolete yesterday” to the
distributed, dynamic organization for our turbulent nowadays
• New ingredients: organizational approaches that enable the utilization of the full potential of our societies
→ Why do we need that? What’s the benefit from social energy?
Case study: Falls Identification• Falls:
• “most significant cause of injury for elderly persons“
• “most serious life-threatening events" in the 65+ age group.
• Technological lock-in• Despite the continuing technological progress, there is no
monitoring system that is able to determine precisely whether a person has actually fallen or, for instance, he or she has changed their position very quickly.
• In some cases, the monitoring system fails.
• Two types of failures:
False negatives and false positives
• False positive:• The system fires an alarm, although the event did not take place.
• False negative:• A fall takes place and the system does not recognize the event as a fall.
Three big problems…
• FN: “In any safety system, false negatives are possibly
the worst kind of failure.”
• Long waiting times: “The single most important factor
influencing the long-term outcome [after a fall] is the
length of time between the fall and getting medical
attention at a hospital. A few hours more or less makes
the difference between life and death.”
(Both quotes: Tom Doris)
• High social costs!
How to improve the service?
• For instance, by using social energy
• Social energy = the power of the people – the use of
society for the sake of improving society’s quality of lives
• A new axiom: from device-only to
→ People as situation identifiers!
People
Devices
AND
The idea (1/2)
• Couple two (or more) fall detectors, F1 and F2
• Both F1 and F2 may be in either the Fallen or NotFallenstate
• Then(F1+F2)(NotFallen) = F1(NotFallen) AND F2(NotFallen)
(F1+F2)(Fallen) = F1(Fallen) OR F2(Fallen)
Less chances to bein NotFallen
More chances to bein Fallen
The idea (2/2)
• People as an extra “detection layer”: when (F1+F2) is in state Fallen, a cloud of volunteers is used to verify if this is
a FP
• Coupling F1 with F2 reduces FN rate
• Using social energy reduces the waste of social resources
• …and reduces waiting times!
• How do we organize this?
• How do we put together?People
Devices
AND
Reasoning & coordination
11
Fractal social organization: building block
Individual &
social concerns
optimization.
Capabilities
Policies
Availability
Location…
Events
PeopleDevices
Bind
Member Member
Member w/
service & feature registry
Service
& featurePublish Publish
+ Communities!
12
Member Member
Service
descriptionPublish Publish
Bind
Local
Reasoning & coordination
Individual &
social concerns
optimization
Capabilities
Policies
Availability
Location…
Events
PeopleDevices MemberMember MemberMember
Exception ⇒ Event propagation
Member w/
service & feature registry
Commu-
nities!
Fractal social organization
An example:
iMinds project
Little Sister
14
Multi-agent simulation
Participating entities:Elderly agents (EA)
Professional carers (PC)
Informal carers(IC)
Device agents (DA)
Mobility agents (MA)
Community agents (CA)
Figure 8. A representation of the FSO structure of the second
simulation model.
NetLogo
Two families of simulations• S1(x): we use only F1 (e.g., an accelerometer) and have a
cloud of x volunteers
• x = 0, 5, 10, …, 40
• S1(0): traditional approach
• S2(x): we use F1+F2 (e.g., accelerometer + gyroscope)
and have a cloud of x volunteers
• Again, x = 0, 5, 10, …, 40
• Each simulation run: 10,000 “ticks”
Coupling F1 + F2: Better FN rate
Social energy: >> handled requests
Social energy: << service costs!
Social energy: << waiting times!
Conclusions
• Challenge: Evolving the organizations while maintaining the
identity of the intended services
• FSO's dynamic hierarchical organization optimally
orchestrates all participating entities thus overcoming the
stiffness of the traditional organizations
• Major returns of social energy:
• improvement of social costs;
• better use of the social resources;
• reduction of the average time to respond to identified falls.
Thank you for your attention!
For more information:
[email protected] / uantwerpen.be
Blog: eraclios.blogspot.be
LinkedIn group "Computational Antifragility"
ANTIFRAGILE workshop
twitter://@EnzoDeFlorio