Credit: CTL

At the 2024 Annual Modeling and Simulation Conference (ANNSIM) in Washington, DC, NIST researchers Serghei Drozdov and Mehdi Dadfarnia gave a 90-minute tutorial on using the NIST-developed open-sourced Python package, SimPROCESD, for the discrete-event simulation of discrete-part, multistage manufacturing systems. This conference attracts experts who showcase cutting edge research in modeling and simulation across various expertise domains.

The SimPROCESD software enables users to rapidly model and simulate part production in any manufacturing configuration that determines the flow of parts through manufacturing machinery and buffer stations to complete a job. The software also allows users to recreate the effects of different maintenance actions, from repairs triggered by Artificial Intelligence (AI)-driven condition-based predictive policies to time-based inspections and run-to-fail corrective work.

The NIST researchers used several examples to showcase the modularity of SimPROCESD's underlying design and its use for production planning and resource scheduling. They also presented a study on using the simulator to understand the risks and benefits of integrating AI-based condition monitoring systems with their maintenance practices. This tutorial represented work from a broader effort in NIST’s Industrial Artificial Intelligence Management and Metrology project, which develops domain-specific tools and methods to improve the effective use of AI systems and tools in industrial applications and to understand their financial and engineering risks and benefits.

SimPROCESD is available for download on Github with accompanying documentation.