Presentation by Paul Valckenaers and Patrick De Maziére during the workshop Interoperability defined by its reason d'être by Paul Valckenaers - AAL Forum 2015
2. Overview
• Introduction:
What is interoperability expected to deliver?
• Interactive:
small groups compile a use case/example.
All (15’) – organisers coach the groups.
• Use case presentations:
Group speakers (3’ each)
• Presentation:
Design for the unexpected and interoperability
• Applying design for the unexpected to use cases:
All (15’) – organisers visit groups and collect info
• End of session (5’).
3. The team
• Paul Valckenaers, UCLL, Paul.Valckenaers@ucll.be
• Bart Haex, KU Leuven, Bart.Haex@kuleuven.be
• Joan Van Loon, IBM, Joan.van.loon@be.ibm.com
• Patrick De Mazière UCLL, Patrick,DeMaziere@ucll.be
4. What is interoperability expected to deliver?
• Commonly assumed / expressed :
“Plug and play”
Achieved by means of a “connector”
Or by complying with a “standard”
• Fact: this is only a matter of (not a lot of) time
provided that:
It is possible (without redeveloping …)
There is sufficient demand (i.e. there is a market)
We enjoy a relatively conflict-free situation
…
5. What is interoperability expected to deliver?
• Implicitly assumed :
Potential of what is available
• Resources (finite, access rights, …)
• …
Coordinated activities, interventions
• Activities
• Interactions through resources
• Orchestrated interactions
• …
6. Interactive session 1
Propose and select a (few simple) use case(s)
from your own experience
Identify the resources, including embedded within …
Identify the services/tools/… to execute
activities on those resources
Assess the accessibility of the resources
Assess the ‘program-ability’ of activities and interactions
Compile a short presentation.
7. Design for the unexpected and interoperability
• Scientific laws or universal principles for the artificial
H. Simon: Flexible (aggregation) hierarchies
Origin of life: Autocatalysis
…
These scientific laws follow from bounded rationality
• Design for the unexpected
Design choices introduce constraints
Repeating design choices builds up inertia
Incompatible constraints cause an inherent
inability to interoperate effectively
• E.g. an interoperability standard incompatible with reality
8. Autocatalysis and the Valley of Death
• Ratio of
User mass
Complexity of the artefact
• Two types of self-reinforcing
Economic ($$$, €€€): paying customers
Information: active users, in a variety of manners
• For all services offering “programmability”
Either (very) simple
Or “mainstream”
Or “decisive features” and …
• In the VoD: many of de jure but not de facto standards
9. Design for the unexpected: Resources
• The real world is coherent and consistent
• Real-world resources …
• Suitable models reflecting real-world resources …
“It is possible”
This always will be “the case” / true
• Ensure/manage access/ownership (also for embedded resources)
Explicit and mandatory resource allocation
• Over time (a reservation service?)
• Accounting for state (transitions)
• Programming features/services stay outside the VoD !!!
Problem domain is stable; user requirements not at all
10. Design for the unexpected:Activities
• Again: real-world reflection > “it is possible”
Preserve the inherent solution space (observe and maximize)
• E.g. robustness in large flexible systems (MIT, axiomatic design)
Manageable solution space for decision-making (must be clever)
• Do not bundle with resources
VIP architecture: activity has a butler making arrangements with …
• Stay out of VoD
Programming services / features
• Interactions
Through resources – provides independance
Proactive (beyond the current discussion)
• SSOT and the D-MAS architectural pattern
11. Interactive session 2
Revisit your results from session 1
What changes? What issues are recognized?
Identify real-world resources and their counterparts in the ICT
domain.
Idem ditto for activities and interactions
How “invariant” are they? Foundation for lasting developments?
Improvements instead of replacements?
Organisers take notes for reporting.
12. Interactive session 2
• Real world resources
Organs
People
…
• Real world activities
Therapy
Travel
…
• Real world interactions
Multi-pharmacy
…
13. Discussion
• Interoperability is more than
Data model
Communication
Standards
• It is about (absence of) constraints
• It is about low inertia (undo) of constraints
• System architecture and architectural patterns
Provide the “it is possible”
Allow reality (the world of interest) to be visible in the IT
• The science behind it (next slide).
14. Design for the Unexpected, 1st Edition
From Holonic Manufacturing Systems towards a
Humane Mechatronics Society
Paul Valckenaers,Hendrik Van Brussel
Expected Release Date: 02 Nov 2015
Imprint: Butterworth-Heinemann
Print Book ISBN : 9780128036624