L-SIM What's New

Summer 2009 - Signal Event

Signal Event is a new event type introduced in BPMN 1.1. It can be used as Start, Intermediate or End event and is denoted as a triangle. Signals offer a general inter-process communications where interested parties can listen for a signal to be raised.

Autumn 2009 - Simulated Annealing Algorithm Optimizer

Simulated Annealing Algorithm Optimizer is an intelligent helper, working tirelessly to improve your results and finds the best solutions for your simulation model. Success is typically measured in terms of service level, throughput or profitability (taking into account cost).

The simulated annealing algorithm (proven in WITNESS) is provided to work as a web service. The algorithm is used to determine the next scenario (set of factor values) based on previous scenarios and objective function results. The Optimizer works with the L-SIM Experimenter service which 'fires off' the L-SIM simulation run, retrieves the responses at the end of the run, and then requests the next scenario from the Optimizer component.

Winter 2009 - User Distributions

User Distributions implement functionality to represent realistic variation in all manner of expressions. User Distributions are for situations where standard distributions do not apply, or if you have real-life data collected from a backend system, you may create your own user defined distribution and use it in a model to take samples from it.

User distributions can be used to create business data e.g. Activity times, Loan values, Passenger type, Exam grade, Resource required. User distributions are constructed from lists of paired data, each data pair has a value and a weight. The sequence of pairs forms a probability distribution rather like a histogram. User distributions are used in very similar way to the existing distributions, i.e. they can be used in any L-SIM expression field. At runtime, the distribution is evaluated by sampling a value from the distribution.

Winter 2009 - Resource Priorities per Task

When modelling resources, organisational structures are assigned and a Priority attribute for each resource is set. Within L-SIM, this has been extended to support different resource priorities for each function (activity) separately. This is useful if several employees are able to carry out a task, but certain employees are better than others at certain functions (activities) due to differing skill sets.

Due New Year 2010 - Governance, Risk and Compliance Simulation

GRC simulation provides answers to the effects of risk occurrence and control execution on process performance and cost. Typical questions are concerned with the impact of risk occurrence on throughput times, bottlenecks or the utilization of resources. These are areas where a dynamical evaluation over time (i.e. simulation) gives additional insight about the system behaviour.

To support GRC, L-SIM is introducing two new object types, a Risk and Control object. The Risk object describes the probability of a risk occurring and specifies the amount of financial damages inflicted when the risk occurs. The Risk object can also trigger an independent process whose responsibility is to clean-up or compensate when a risk is detected. The Control object helps to either reduce the probability of the Risk occurrence or it reduces the damage caused by the Risk.

This capability is being developed to support our partner IDS Scheer; further details of their plans can be found at http://www.ariscommunity.com/users/elba/2009-11-19-process-risk-simulation-soon-come