Quasi-real time computing and combination of temporal context information with spatial and attributive context information in dynamically changing situations and processes are of specific interest.
There is a rule-based modelling language and inference engine that allows to model temporal dependencies between contexts, provides mechanism for modeling and processing uncertain context, and allows for basic statistical analysis of past context: http://glados.kis.agh.edu.pl/doku.php?id=pub:software:heartdroid:start
If not this, Dynamic Bayesian networks, or Markov Models could be used to handle your problem as well I guess.
In the context of smart/interactive materials which have dynamic/ temporal qualities, there is a recently developed 'simulator' - as a tool for design professionals, by the PhD candidate Bahar Barati.