Comparing supply chain optimization and simulation
Dan Gilmore has written an article at Supply Chain Digest comparing the merits of supply chain optimization versus simulation.
In my mind, they are simply two approaches suited for two different problems. Years ago the focus of all supply chains was efficiency and trying to run things like clockwork. Process automation and removing people from the process were key objectives. Today, managing volatility and change is critical to success, so the more urgent need has become how to respond, and that requires human judgment. So, where optimization can be used when assumptions are known and stable, simulation that supports human judgment is required when there is frequent change requiring rapid scenario analysis to determine the right course corrections to make.
Steve Hochman of AMR Research just published a paper entitled “Response Management: Next Wave in Supply Chain Innovation?” talking precisely about this issue (get the paper here). In the paper, Steve says “Our research shows conventional technology hasn’t met the need. Consider the following:
Advanced planning–Automated algorithms are tailor made for low-value-intensity, high-complexity decisions. But collective judgment becomes an essential complement in context-rich decisions (for example, will this irate customer defect if I don’t pay $500,000 in expediting costs?).
Spreadsheets–Standalone spreadsheets are pervasive precisely because of the limitations of closed-loop planning. The problem, of course, is that spreadsheet flexibility quickly devolves into multiple versions of the truth.
Business intelligence (BI)–As with spreadsheets, BI dashboards put decision control back in the hands of people. But BI tools are structured for insight, not execution. In the face of difficult real-time tradeoffs, yesterday’s snapshot of read-only analytics leaves a wide gap between awareness and confident action.
In environments where decisions are high in risk and complexity, many find they need systems for collective risk tradeoff and response. Whether the risk takes the form of a forecast spike, a component approaching obsolescence, a sales order shortfall, or more strategic sales and operations planning (S&OP) or network commitments, discrete manufacturing clients increasingly place a premium on the ability to see, study, and simulate alternate resolutions before pulling the decision trigger on a major tradeoff.”
So, I don’t think it’s an either/or situation, both have their place - although I would argue the more urgent need today is for staff empowerment to deal with the unexpected, and simulation is a core tool to accomplish this.
