New Directions in Modelling Innovation: applying agent

New Directions in Modelling Innovation:
applying agent-based methods to
socio-technological transitions
Jonathan Köhler, Michaela Gigli, Antje Bierwisch, Arne Lüllmann
Fraunhofer ISI
[email protected]
EU-SPRI Conference 2012
Fraunhofer ISI
Karlsruhe
12-13th June 2012
Introduction
Transition Theory
Agent-Based modelling of transitions: the MATISSE pilot model
Psychology of behavioural change
Niches and regimes: agents, networks, strategies
Possible modelling methods
What is to be done?
Increasing structuration
of activities in local practices
Transition Theory
Socio-technical’
landscape
(exogenous
context)
Multi-level perspective
on transitions
(adapted from Geels,
2002: p. 1263)
Landscape developments
put pressure on existing regime,
which opens up,
creating windows
of opportunity for novelties
Markets, user
preferences
Sociotechnical
regime
New regime
influences
landscape
Industry
Science
Policy
Culture
Technology
Socio-technical regime is ‘dynamically stable’.
On different dimensions there are ongoing processes
External influences on niches
(via expectations and networks)
New configuration breaks through, taking
advantage of ‘windows of opportunity’.
Adjustments occur in socio-technical regime.
Elements become aligned,
and stabilise in a dominant design.
Internal momentum increases.
Nicheinnovations
Small networks of actors support novelties on the basis of expectations and visions.
Learning processes take place on multiple dimensions (co-construction).
Efforts to link different elements in a seamless web.
Time
Agent-Based modelling of transitions: the MATISSE
pilot model:1
MACRO
SUPPORT
Complex agents
MICRO
Empowered
Niche Agent
Simple
Agents
NICHES
CONSUMERS
Changes in Preferences, Practices
REGIME
MESO
Effectiveness in generating
strength from support
LANDSCAPE
Agent-Based modelling of transitions: the MATISSE
pilot model:2
Two illustrations of a two-dimensional practices space, with practice axes Px and Py.
Left: regime and niches, which can move in the space and interact with each other.
Right: the consumer agents showing supporters scattered in the practices space,
coloured by the agent they support, red = regime (R), green = niche 1 (N1), blue = niche
2 (N2).
Py
PY
N2
Niche 2
Regime
PX
Px
Niche 1
R
N1
Psychology of behavioural change
Heuristic framework model (adapted from Prochnow et al, 2012)
Psychology of behavioural change
Levels and theories of behaviour change: electric vehicle adoption.
Anable et al. (2011, p.1017)
Niches and regimes: agents, networks, strategies
Niches, regimes:
Model separately or emergent property from agent interactions?
Can they be modelled as networks, using e.g. social network theory?
Functioning of niche markets:
Firm strategies; strategic managment ideas (as opposed to Strategic
Niche Management)
Social networks and entrepreneurial activity, network dynamics
Integration of product based characteristics and consumer characteristics in
Agent Based Models
Possible modelling methods
Problems of calibration with large number of parameters:
the inverse problem, from overall result in the data, identify
ranges of parameters
Game theory for agent interactions
Neural nets, learning algorithms from Artificial Intelligence
plausible
What is to be done?
Research question:
how to develop an Agent Based Model of sustainability transitions that can
investigate how niches arise and grow and under what conditions they lead to a
transtion to a new socio-technical regime.
current literature on modelling does not address behavioural changes towards
sustainability and niche growth processes
Possible methods
Psychological models for agents (consumers)
Social structures of niches and regimes, social network ideas
Firms’ behaviours and the dynamics of niche structures on the supply side,
Strategic management theory as well as Strategic Niche Management