Manufacturing Discrete-Event Process Optimization
Budget: ₹12,500 – ₹37,500 INR
I need a discrete-event model built in AnyLogic that captures a real manufacturing line, right down to individual machines, buffers, shift patterns, and change-over logic. The objective is clear: identify bottlenecks and test improvement ideas so I can decide which mix of resources, sequencing rules, or layout tweaks will push throughput to its practical ceiling while keeping WIP and cycle time in check.
You will start from process maps, cycle-time studies, and production data I provide (CSV/Excel). Once the flow logic is reproduced faithfully, we will iterate through validation runs until the model’s KPIs match historical performance within an agreed tolerance. After that, you’ll configure experiments for scenario comparison and built-in optimisation so I can explore “what-if” questions without touching the model code myself.
Deliverables
• An AnyLogic .alp file with clean, well-commented structure
• Parameter sets and experiments covering the baseline plus at least three optimisation scenarios
• A brief slide deck or PDF summarising model assumptions, verification results, and key insights
Acceptance criteria: model output aligns with actual line data (±5 % across throughput and average queue length) and runs fast enough to complete 100 replications of a one-shift day in under five minutes on a standard laptop.
Throughout, please keep all source files, data links, and documentation organised so I can maintain and extend the model after project hand-off.
You will start from process maps, cycle-time studies, and production data I provide (CSV/Excel). Once the flow logic is reproduced faithfully, we will iterate through validation runs until the model’s KPIs match historical performance within an agreed tolerance. After that, you’ll configure experiments for scenario comparison and built-in optimisation so I can explore “what-if” questions without touching the model code myself.
Deliverables
• An AnyLogic .alp file with clean, well-commented structure
• Parameter sets and experiments covering the baseline plus at least three optimisation scenarios
• A brief slide deck or PDF summarising model assumptions, verification results, and key insights
Acceptance criteria: model output aligns with actual line data (±5 % across throughput and average queue length) and runs fast enough to complete 100 replications of a one-shift day in under five minutes on a standard laptop.
Throughout, please keep all source files, data links, and documentation organised so I can maintain and extend the model after project hand-off.