Abstract for: A Technique for Estimating the Progress of System Dynamics

System dynamics has evolved from a qualitative, “servo” approach to a quantitative, data-based, predictive capability that resembles the control theory field of Model Predictive Control (MPC). We provide and apply a technique for estimating the progress of SD in MPC mode. We have created a simulation test bed to quantify how much this transition has improved the value of SD in improving the performance of systems. The test bed consits of a Python codebase that allows us to run two instances of a model in parallel and analyze how a model run in MPC mode performs relative to a model run in servo mode. Results suggests that the improvement may have been large. Because the MPC use of SD models relies on the predictive power of SD models, we also tested the sensitivity of the results to errors in prediction, and discovered that the MPC use of SD models has high payoff even if the predictions are imprecise. Our preliminary results suggest serious improvements are possible if organizations would imbed SD models in ongoing decision streams, as in MPC. We hope to see organizations grow to take advantage of the full capabilities of system dynamics models, with resulting major improvements in performance. For reviewing, summarizing and condensing human-generated text