What is a plan

A Plan is a projection of future developments is the design of future changes 
A Plan represents how we will or we should meet current and future needs

A Plan describes decisions to be implemented 

We focus on spatial decisions and hence spatial plans: architectural plans and urban plans
(e.g. building layout, building form, land use plan, facility location-allocation) 
Plans are not always good representations of our future needs but are always part of the dynamic of change in cities. 

What is artificial plan designing

Artificial plan designing refers to the use of computational models to generate plan descriptions. 
Sometimes it  is referred to as “automatic plan generation” or “creative design”.

Artificial plan designing  is designing:

by machines and computational models 
by human-machine interaction 
by human- machine networks interaction 

What is the motivation for artificial plan designing research 

Designing and planning are cyclic processes. 
Artificial Plan Design supports the generation of solutions and evaluations all the way through the different stages of design.

Designing and planning are collective and multidisciplinary processes. They require the participation of multiple agents  with diverse requirements and views. 
Artificial Plan Design can be used to explore diverse requirements and interdependencies in order to support collective decision making and conflict resolution.

Design and planning are creative processes.
Artificial Plan Design makes all trivial computations for us. But it may also reveal unforeseen solutions.

Some typical methodological approaches 

Optimization:considers design as a search for optimum solutions- uses various optimization techniques to find best solutions under certain constraints. 

Evolutionary Design:considers design as a selective and creative process-  uses genetic algorithms to  create evolving design forms.  

Shape Grammars: considers design as a compositional process based on shape vocabularies -  generates designs based on the creation  and reproduction of grammatical rules. 

Case-Based Reasoning: considers design as a creative process- focuses on the adaptation and reuse of previous knowledge incorporated in design cases to solve similar problems.

Multi-Agent Coordination: considers design as a search for coordinated solutions that satisfy distributed domain requirements and views.    

A model for artificial plan designing by “learning control coordination”

We propose a model-tool that learns by user interaction and then uses this knowledge to generate plan designs 

The problem is defined as a search for locations and physical layout proposals that satisfy distributed and time-variant targets. Expert knowledge for this search is not explicitly incorporated in the model but a Neural Network is used instead to discover and represent interdependencies among decision variables expressed by distributed sources (decision makers and their domain models.

We present a model-tool that generates plan descriptions based on two basic functions:
A Neural Network model  learns by agent interaction associations among decision variables (e.g. associations between building form and energy consumption) 
Then the model uses this knowledge as a guideline to produce solutions that satisfy distributed targets.

The design problem as multi-agent coordination

We consider that plan descriptions are built on distributed domain problems and partial proposals developed by agents (human or artificial). For this model plan descriptions are generated within a virtual reality (VR) world and are composed by aggregated objects introduced by users. The specifications of these objects are dynamically identified and modified by human actors directly or via the use of formal models.
The design problem is formulated as a coordination problem among self-interest agents (which are represented as cubes in the VR world) and is addressed via a distributed learning control methodology. 
Plan descriptions represent coordinated solutions (changes) that satisfy distributed domain requirements and views.

Control based coordination method

The purpose of design based on the control methodology is to find solutions that can direct partial descriptions produced by agents, to follow their dynamically defined targets -despite conflicting requirements.
Control is the process of self-adaptation of agents that leads to coordination of their distributed description.

Experimentation

A control model has been built based on Adaptive Backthrough Control architectures.
These structures typically use two neural networks: the Controller (the system that controls) and the Plant Model (a model of the system to be controlled). First, the plant model is trained to approximate the plant by learning on line or off-line input-output patterns of the agent behaviour. Then, these patterns are used “backwards” as a guideline for the controller.
   
We have experimented with mathematical formulations that model agent behaviour (like motion, shape transformation and costs) based on state space methodology, as well as with fuzzy systems.