Schematic plans
Power& tradition holding back progress Contents Overcome reactionaries

Schematic planning

Summary
This presentation reviews planning using complex adaptive system (CAS) theory.  Genes and ideas reflect CAS properties. 
Introduction
Some of us struggle to limit our alcohol consumption or our over-eating.  But with practice our unconcious minds can learn to play musical instruments and drive cars effectively.  Why is that?  This presentation:
  • Reviews how emergent(p.21) systems can represent and benefit from state(p.3)
  • Describes key problems(p.4) inherent in effective planning in uncertain situations
  • Introduces a schematic structure(p.12) used by evolved systems to overcome these challenges and operate effectively in their proximate environments
  • Outlines the interaction of schematic structures with agents(p.13) and genetic operators(p.15)
  • Reviews in detail a document set(p.18) which
    • Encourages flexibility
    • Supports institutionalized learning
    • Enables change
  • Lists opportunities and challenges(p.19) to introducing these techniques in companywide business planning processes

- next page -> (1 of 19)
 

Power& tradition holding back progress Contents Overcome reactionaries
rhlcontents

<- previous page - - next page -> (2 of 19)
 

Power& tradition holding back progress Contents Overcome reactionaries
Plans of great flexibility can be represented using a schematic structure(p.12) 

Molecular genetics and brains depend on this form of representation

Opportunity:
  • Emergent systems including businesses and biological systems need to represent their situation and the various strategies they can deploy to survive
  • Biological systems all use the same basic structures, 'genes', to do this
    • Grow, learn and compete by leveraging this genetic platform
  • World wide web's HTTP and HTML technologies provide facilities for developing analogous schematic structures
  • Businesses can benefit by implementing evolution's strategies on web based schematic structures enabling
    • Enhanced organizational awareness
    • Learning
    • Creativity


<- previous page - - next page -> (3 of 19)
 

Power& tradition holding back progress Contents Overcome reactionaries
The planning problem:
  • Generalized awareness is difficult to develop - Higher animals need extensive networks of cellular 'agents' to maintain effective models of their environments
    • They are fooled by ambiguous situations (illusions)
    • They must prioritize between different aspects of the situation and what to do
    • Strategies of explicit design have been
      • Associated with specific situations, and
      • Do not by themselves support the flexibility to process general environments
    • Evolved strategies are constrained by emergence and focused by the proximate environment
      • Rapidly changing environments and increased connectedness present problems and opportunities 
      • Identifying where agents need help and how to provide it should be part of the plans
    • Environmental boundary is difficult to define
      • Emergent agents(p.13) with replicated plans can aggregate
        • They can operate as cooperating multi-agent systems
      • Multiple schematic networks may be present within a multi-agent system
        • Some of these may depend on histories represented in lower levels of emergence
      • Laws may validly hold for a higher level emergent agent
        • At lower levels the laws are only reflected in transient structures driving the strategies of agents


<- previous page - - next page -> (4 of 19)
 

Power& tradition holding back progress Contents Overcome reactionaries
The planning problem (continued - models):
  • Effective learning in a situation of uncertainty requires a modeling(p.14) method and the capabilities to perform it
    • They must identify which models apply to the situation they are experiencing
      • It may require luck to avoid selecting a model which leads to a trap or confusion
      • Evolution uses parallel agents with differing plans and state to overcome this problem through selection
    • Hypotheses are used as assertions about the situation until they are shown to be false
    • Sensors that respond to particular aspects of the environment can provide signals about the situation
    • Actions are performed by agents(p.13) in response to particular signal clusters from the environment
    • Alternative hypothesis about the situation can be integrated into models which test the current situation and indicate how close a match each hypothesis is
    • Strategies associated with each identified scenario are then used to frame the situation
      • Strategy associated with the most likely scenario is acted upon
  • Environment includes other adaptive entities
    • Other entities strategies and actions must also be modeled and represented


<- previous page - - next page -> (5 of 19)
 

Power& tradition holding back progress Contents Overcome reactionaries
The planning problem (continued - extrapolation):
  • Increasing connectedness and interdependency of agents(p.13) increases complexity
    • Historically successful strategies may fail or become counterproductive
    • System flows will transform over time.  Sub populations of agents will be effected asymmetrically by the trends. 
    • Constraints and resource levels will alter.  
  • Only recently we have developed the capability to significantly transform our environment
    • Billions of humans and our animals transform local resources
    • Tools and chemistry enable us to focus large forces at the environment
    • Human evolution has not had time to introduce adaptations that reflect the current environment
  • We have had to cope with catastrophic events but now we can cause them and they are hard for us to recognize and manage
  • 'Innate' spatial representations and models of facial expressions, group focus etc. are of limited use in predicting transforms through time
  • It feels right to prioritize the immediate over the chronic even as our aggregate activities have undermined acute problems and empowered chronic ones
 

<- previous page - - next page -> (6 of 19)
 

Power& tradition holding back progress Contents Overcome reactionaries
The planning problem (continued - plans):
  • Plans are both recipes of how to achieve actions and constraints on what can be done 
    • Specific recipes describe the cascade of activities to achieve a complex goal
    • Various agents(p.13) may be involved in the cascade and may have to coordinate when to act
    • Recipes must be shared between the cooperating agents
  • Our cognitive functions expect recipes to be sequences but many systems are inherently parallel
    • Our cells use a highly parallel chemical environment and the DNA based plans reflect this 
    • Parallel architectures like Google's search engine provide opportunities to cope with and leverage real world complexity
  • Goals can constrain the agents assigned to carry them out to attempt the impossible
    • Preconditions must be used to ensure that the agreement is achievable
    • Failure to achieve dependent tasks should terminate the contingent agreement
    • Artisan personalities resist adopting plans to improve their flexibility
  • Identifying the scope of a plan to manage a complex system is difficult
    • Too much detail can destroy the required focus
    • Too little detail can miss significant aspects or generate actions which are not specific


<- previous page - - next page -> (7 of 19)
 

Power& tradition holding back progress Contents Overcome reactionaries
The planning problem (continued - doing):
  • Executing the actions defined in the plan takes time and resources
  • While executing actions agents(p.13) get close to the operations and focused on achieving them. 
    • Well trained agents can use their highly evolved senses to spot problems with the resources and processes
    • Shewhart cycles(p.22) allow the issues to be corrected and re-planned


<- previous page - - next page -> (8 of 19)
 

Power& tradition holding back progress Contents Overcome reactionaries
The planning problem (continued - checking):
  • Evaluating the success of actions takes time and failure creates doubts and undermines value
    • Success is uncertain
    • Future is unpredictable
    • Intermediate actions may be necessary but appear relatively ineffective
    • Models(p.14) must become aware of the situation specific valuable associations
    • So many agents assume success
      • Evolved strategy - so it has advantages to the selfish genes
      • Gaining time and possibly status 
      • Looking for scapegoats when a failure is detected
  • Fitness functions provide immediate feedback and allow separation of the problem into two well-defined phases used by genetic algorithms(p.15):
    1. Operation of agents(p.13) to determine aggregate fitness based on the fitness functions
    2. Generation of a next generation of schematic structures(p.12) based on current populations aggregate fitness
  • More typically the persistence of particular agents to survive long enough to reproduce and support its offspring to maturity is not described effectively with simple fitness functions
    • They must explore their proximate environment creatively to
      • Identify new niches
      • Discover/develop additional sensors, models and actions
      • Conserve these creative capabilities
      • Capture resources and use them to enable eventual reproduction


<- previous page - - next page -> (9 of 19)
 

Power& tradition holding back progress Contents Overcome reactionaries
The planning problem (continued - corrective actions):
  • Planning in uncertain environments aims to capture and retain structures which maximize success in the proximate environment
  • Requires agents(p.13) to dedicate time between:
    1. Exploring the features of the proximate environment, using current strategies, to expand knowledge of available options - features and strategies may change
      • Huge search space requires that current structural state is used to focus the search - Operational structures should be based on broadly valued current genetic sets
      • Use of massively parallel searching
    2. Identifying structures which generate valuable associations of the various tested situations and current agent strategies, which include genetic operators(p.15)
      • Parallel operations helpful
      • Emergent strategies must be based on the structures and so will change


<- previous page - - next page -> (10 of 19)
 

Power& tradition holding back progress Contents Overcome reactionaries
The planning problem (continued - selection and genetic operations):
  • Tests must contribute to a networked history representing the fittest sample of goals, strategies, models(p.14), and recipes(p.12) for the proximate environments encountered
    • History will represent high value reproductive (germ-line) plans, and operational (somatic) instantiations of them
      • By presenting a broad range of situations during testing of a particular generation of competing agents(p.13) plans can become robust across a range of situations
      • In general the result is a network, not the hierarchy which Hoshin advocated
        • Means coordination of the genetic structures is complex
        • Integration process must be architected
  • Planners then aim to identify both the niches where the business will operate and the fittest of the samples to execute
    • Slight variations (mutations) are introduced which are viewed as an operation's unique added value
      • Iterative Shewhart cycles(p.22) allow the planners to validate their assumptions and make corrections


<- previous page - - next page -> (11 of 19)
 

Power& tradition holding back progress Contents Overcome reactionaries
Introduction to schematic structures
Schematic structures are constructed from simple building blocks, such as nucleic acids or the letters of the alphabet.  These blocks are able to link together into groups, lists and networks, and directly or indirectly associate with agents(p.13)

  • Deployed based on a fundamental building blocks (1) which can: schematic string
    • Provide brick for agents, including genetic operators(p.15), to build structures
    • Represent a data value and be
    • Replicated and
    • Linked into lists (2)
    • Integrate descriptors, and addressable labels
      • Descriptors can be interpreted broadly by agents as for example valuations or epi-genetic control signals
    • Associatively network
  • Any schematic sequence of basic elements can be interpreted by an agent as a signal (3) to perform a specific act
    • Agents position on the schematic structure represents a state and so the local sequence can be a signal
    • Becomes an association (4) between signalled event and the agent's action
    • Actions can include modeling(p.14) of a situation


<- previous page - - next page -> (12 of 19)
 

Power& tradition holding back progress Contents Overcome reactionaries
Emergent agents process environmental signals and their instance of the schematic plan to select actions which they perform
Agents can interpret the schematic structures as they please, but acted on by evolutionary selection beneficial associations of signals and actions transform into strategic attributes of the system
  • A labeled element, operon, with associated actions corresponds to a goal
  • Specific goals can be associated with multiple, alternative actions
  • Individual agents can be deployed with a full complement of operons, but different, allelic, sets of associated actions
  • Filtering of labels and other tags provides a flexible control of flows
    • Agents can filter, and integrate the flows of schematic elements to perform genetic operations(p.15)
  • Genetic operations provide the basis for learning, by correlating success in replication of schematic structures and the retention of valuable sequences with positive agent actions
    • Requires that agents actions are based on schematic signals
  • Total set of different germ-line schematic structures shared in genetic operations corresponds to the evolving state
  • Ability to replicate schematic structures allows for somatic copies to be augmented with highly specific state descriptors, while germ-line copies are left pristine
  • Multi-cellular agents are able to deploy different states in individual somatic cells


<- previous page - - next page -> (13 of 19)
 

Power& tradition holding back progress Contents Overcome reactionaries
Models provide agents with a schematic view of the world
  • Emergent agents(p.13) must be able to differentiate good strategies from bad.  Models assist them with making such decisions
  • They are hypothesis that use success or failure as an indicator of legitimacy
  • Correlation of model and schematic(p.12) agent action allows evolutionary selection to 'rate' models in particular niche situations
  • Agents can extend the evolved rating mechanism by applying a Shewhart cycle(p.22)


<- previous page - - next page -> (14 of 19)
 

Power& tradition holding back progress Contents Overcome reactionaries
Modifications to the germ-line structures are performed by genetic operators 
  • Emergent agents(p.13) can include genetic operators which allow reproduction to formally transform the current structures for the next generation
  • Genetic operators can equivalently be deployed via a discrete process called a genetic algorithm
  • Operations include: mutation, crossover, inversion, translocation, deletion, dominance modification


<- previous page - - next page -> (15 of 19)
 

Power& tradition holding back progress Contents Overcome reactionaries
Uncertainty about the situation requires situation-specific hypothesis to be represented and the situation to be perceived as a set of correlated environmental signals
  • Networks of associations allow the representation of alternative hypothesis for each historically experienced situation (local signal set) along with evaluations of each
  • Initial schematic structures generated by replication include recipes that agents(p.13) can follow to deploy sensors for historically useful signals
  • Advanced somatic perception and representation architectures, such as brains, provide mechanisms to integrate new structures over time as their value becomes perceived as consistently high
    • Initially new structures are indirectly referenced, and may be dereferenced if their value is limited and they conflict with other situationally active structures
    • If the new structures contribute valuably, and become extensively referenced by other active structures their associations can be made long term


<- previous page - - next page -> (16 of 19)
 

Power& tradition holding back progress Contents Overcome reactionaries
Schematic representations based on web protocols
  • HTML allows a linear schematic representation.  It provides linkages between name (id) and href to allow text structures to be associated
  • HTTP allows schematic structures to be linked into a network
  • Germ-line and somatic schematic web document networks provide businesses with the same benefits and challenges as biological systems
  • Genetic operations(p.15) can leverage niche diversity, but must value the additional action, or model(p.14), structures associated with an allelic goal
  • Network structures and communications can support deployment and use of perception and representation structures and state


<- previous page - - next page -> (17 of 19)
 

Power& tradition holding back progress Contents Overcome reactionaries
An illustrative network of schematic structures
  • Reference schemata structured as a list of hypothesis(p.4) areas with links to each valued hypothesis, and its associated test results and situational values
  • Situational analysis structured as a list of situation areas associatively linked to the situations
    • Each situation is linked associatively to tests, and flags associated with alternative hypothetical  scenarios
  • Other agent(p.9) models(p.14) structured as a list of agent areas associatively linked to each agent
    • Each agent is assumed to use similar models and strategies to our own
  • Strategic analysis structured as a list of analysis areas associatively linked to each analysis
    • Each analysis is associatively linked to operational goals, and to trusted reference schemata
  • Operational plan is structured as a list of goals associatively linked to action sets
    • Each goal has an adjustable priority indicating how salient its actions currently are
    • Each action has an associated agent who will perform the action if it is sufficiently salient
  • Value network described as a list of nodes and linkages
    • Each node's properties is defined


<- previous page - - next page -> (18 of 19)
 

Power& tradition holding back progress Contents Overcome reactionaries
Schematic plans offer the potential to:
  • Enable cascading goals to be formally associated with operational activities, their priorities, status and ownership
  • Correlate analysis with goals and strategic alternatives
  • Improve our organizational learning with effective processes defined and integrated into the core knowledge of the business
  • Highlight the impact of education in the operation of organizations
  • Leverage the evolutionary nature of organizational change
  • Encourage the development of formal integration of the plans and viewpoints of operations with those of corporate executives
While gaining a better understanding of the equivalent nature of the challenge for, and elegance of, our brains

There are challenges:
  • Need for processes to integrate operational schematic documents with the corporate masters will highlight misalignments and divergent perspectives
  • Providing business model niches for breakthrough businesses which cannot align with the major organizations will be necessary and will require vision and determined executive sponsors
In overcoming these challenges business planning should provide increased value

 


<- previous page - (19)
 

Power& tradition holding back progress Contents Overcome reactionaries
Backup slides follow

- next page -> (backup 20 of 22)
 

Power& tradition holding back progress Contents Overcome reactionaries
Emergence, adaptation and complexity 
The US sub-prime housing bubble and crash, the rapidly increasing cost of US health care, and the development of coral atolls are each good examples of effects of a complex adaptive system (CAS).  In each case the system had many participants(p.13) interacting and responding to each other.  The participants in each system:
  • Adapt to their situation in complex ways, using a multiplicity of strategies and actions based on capabilities they were born with, and techniques they have learned,
    • The actions of each participant is interdependent
    • Changing one aspect of the system will alter other parts. 
    • Correctly predicting the result requires an accurate model(p.14) of the changes that will occur in each of the effected participants
  • Striving to capture resources: credit, effective treatments, or light, food and shelter, so they can
  • Continue with their lives and raise offspring
  • Each of the participants in the housing and health care networks is an aggregate of components which also respond to their local situation based on a recipe of action
  • Each participant uses one or more recipe style plans to define their potential actions
  • The participants, and their building blocks, repeatedly adopt and respond to strategies e.g. in housing and health care including: flexibility, centralization, prophylaxis, infrastructure amplifier, evolved amplifier, platform, and disruption


<- previous page - - next page -> (backup 21 of 22)
 

Power& tradition holding back progress Contents Overcome reactionaries
Shewhart cycles: Walter Shewhart advocated the use of iterative cycles of: Planning, Doing, Checking and Acting to fix issues (PDCA)
  PDCA cycle emerges naturally in complex adaptive systems
  • Examples include the cell cycle, cell death programme, cell growth programme, Just-In-Time manufacturing
  • Plans in an adaptive system are schematic structures(p.12)
    • Associated with current environmental state indirectly through modeling(p.14)
    • Models generate predictions of results of performing each strategy
  • Performing, or 'do'ing, a highly valued strategy will result in some change of state of the agent(p.13) and its environment
  • Agent can compare (check) the result against the models prediction
    • Human agents evolved strategies often avoid checking(p.9) so a required process supports the formation of the habit
  • If reality differs from the prediction the valuation of the strategy can be adjusted

<- previous page - (backup 22)