This document discusses problems that may arise from interconnected complex adaptive systems in smart cities. It raises questions about what could happen when different systems like smart homes, mobility networks, and the power grid interact with each other. Some specific questions asked include: how can these systems account for human irrationality; how can we trust systems that are prone to emergent behaviors; and how can we design systems to be self-explaining when failures occur due to the interaction of independent parts. The document uses a hypothetical example of a lunch plan being disrupted by public transport strikes to illustrate issues that could arise from unforeseen emergent behaviors in interconnected smart city systems.
1 of 25
Download to read offline
More Related Content
Where Shall We Have Lunch? Problems For A Computer-aided Future
1. CONTEXT / BACKGROUND THE PROBLEM THE BIG QUESTIONS
Where shall we have lunch?
Problems for a computer-aided future
Vivek Nallur
Trinity College Dublin
July 20, 2015
2. CONTEXT / BACKGROUND THE PROBLEM THE BIG QUESTIONS
THE BIG QUESTIONS
The History of every major Galactic Civilization tends to pass
through three distinct and recognizable phases, those of
Survival, Inquiry and Sophistication, otherwise known as the
How, Why, and Where phases. For instance, the 鍖rst phase is
characterized by the question How can we eat? the second by
the question Why do we eat? and the third by the question
Where shall we have lunch?
Douglas Adams, The Restaurant at the End of the Universe
8. CONTEXT / BACKGROUND THE PROBLEM THE BIG QUESTIONS
SO, WHATS THE PROBLEM?
Figure: Visualization of movement on the LondonUnderground
9. CONTEXT / BACKGROUND THE PROBLEM THE BIG QUESTIONS
EACH HOME IS A COMPLEX ADAPTIVE SYSTEM
Figure: https://commons.wikimedia.org/wiki/File:
Complex-adaptive-system.jpg
10. CONTEXT / BACKGROUND THE PROBLEM THE BIG QUESTIONS
IF DUBLIN WERE A SMART CITY
Number of smart homes 400,000 (approx)
Number of smart buses 950 (approx)
Number of ansibles 1 million (approx)
Number of Luas riders per day 83,500 (approx)
Number of Bus riders per day 300, 000 (approx)
How many feedback loops does this system have?
11. CONTEXT / BACKGROUND THE PROBLEM THE BIG QUESTIONS
LETS HAVE LUNCH!
My ansible can talk to your ansible
Coordinate schedules
Contact city transport network for options, pre-and-post
lunch
Find restaurants amenable to both palettes and budgets
Make reservations
Reserve place on city transport vehicles
What could possibly go wrong?
12. CONTEXT / BACKGROUND THE PROBLEM THE BIG QUESTIONS
STRIKE! STRIKE! STRIKE!
Talks breakdown between management and union
Union announces demonstrations, sit-ins, protests
Self-Adaptation By Smart Transport Network
Re-route buses, trams
Change schedules to accommodate rush
Cancel reservations on public transport network, lying on
strike route
13. CONTEXT / BACKGROUND THE PROBLEM THE BIG QUESTIONS
Self-Adaptation By Smart Grid
Observe spike in power demand from tram-network
Re-route power from residences to tram-network
Start-up generators to cope with future demand in
residences
Self-Adaptation By Smart Home
Change schedules of automated washing machines,
dishwashers
All non-essential appliances switched off
Self-Adaptation By Ansibles
Start charging EVs for un-missable appointments
Cancel appointments that cannot be met
Change schedules
14. CONTEXT / BACKGROUND THE PROBLEM THE BIG QUESTIONS
WHATS WRONG WITH ALL OF THAT?
What happened to my lunch appointment?
The ansible has no idea
The smart-home has no idea
The smart-grid has no idea
Series of cascading events
Disruption in schedules cause changes to traf鍖c patterns
Changed traf鍖c patterns cause more power to be drawn by
transport networks
Sudden power demand causes brownouts in residential
neighbourhoods
15. CONTEXT / BACKGROUND THE PROBLEM THE BIG QUESTIONS
AND YOURE JUST ONE IN A MILLION CITIZENS
16. CONTEXT / BACKGROUND THE PROBLEM THE BIG QUESTIONS
WAS THERE AN ACTUAL STRIKE?
Can you trust an autonomous complex adaptive system?
Which agent is responsible for failure?
Can the system explain itself?
What about emergent phenomena?
Reasoning about failure
When no single system has failed
And no-one has the big picture
17. CONTEXT / BACKGROUND THE PROBLEM THE BIG QUESTIONS
ALREADY COMPLEX SYSTEMS ARE NERVE-WRACKING
In Stock-Exchanges
Flash crashes caused by high-frequency trading algorithms
60 - 70% of price movements caused by self-generated
activity
Oct, 2014 saw 179 mini-鍖ash-crashes
Air-Traf鍖c Control
December, 2014 ATC shutdown airspace over London
18. CONTEXT / BACKGROUND THE PROBLEM THE BIG QUESTIONS
INTER-CONNECTED COMPLEX SYSTEMS
Need Self-Explaining Mechanisms
To account for humans in the loop
To account to humans in the loop
Will almost certainly experience Emergence
19. CONTEXT / BACKGROUND THE PROBLEM THE BIG QUESTIONS
QUESTIONS THAT WE DESPERATELY NEED ANSWERS
TO
What happens when one complex adaptive system meets
another?
Smart Homes meets Smart Mobility meets Smart Grid
20. CONTEXT / BACKGROUND THE PROBLEM THE BIG QUESTIONS
QUESTIONS THAT WE DESPERATELY NEED ANSWERS
TO
What happens when one complex adaptive system meets
another?
Smart Homes meets Smart Mobility meets Smart Grid
How do we account for irrational human behaviour?
I only eat vegetarian on Wednesdays, unless Lorettas has saltimbocca
alla Romana as the lunch special
21. CONTEXT / BACKGROUND THE PROBLEM THE BIG QUESTIONS
QUESTIONS THAT WE DESPERATELY NEED ANSWERS
TO
What happens when one complex adaptive system meets
another?
Smart Homes meets Smart Mobility meets Smart Grid
How do we account for irrational human behaviour?
I only eat vegetarian on Wednesdays, unless Lorettas has saltimbocca
alla Romana as the lunch special
How far can we trust a socio-technical complex adaptive
system?
If my ansible says not to take the Waterloo line, then there must be a
fault
22. CONTEXT / BACKGROUND THE PROBLEM THE BIG QUESTIONS
MORE TECHNICAL QUESTIONS ...
How do we construct a socio-technical complex adaptive
system?
23. CONTEXT / BACKGROUND THE PROBLEM THE BIG QUESTIONS
MORE TECHNICAL QUESTIONS ...
How do we construct a socio-technical complex adaptive
system?
What kind of design philosophies exist for such systems?
24. CONTEXT / BACKGROUND THE PROBLEM THE BIG QUESTIONS
MORE TECHNICAL QUESTIONS ...
How do we construct a socio-technical complex adaptive
system?
What kind of design philosophies exist for such systems?
If emergence is a given, what are the implications for
proof?
25. CONTEXT / BACKGROUND THE PROBLEM THE BIG QUESTIONS
THATS ALL, FOLKS!
Questions? Comments?