Tierwise

What it does

• Six tiers, several suppliers per tierConsumption points, warehouses, packaging CMOs, final dosing CMOs, product manufacturing CMOs and raw ingredient CMOs. Each site sources from one or more sites in the tier above, by share.
• Latitudes and seasonsSites run from 52°N to 38°S. Demand peaks in the local cold season, six months apart between hemispheres. Raw milk capacity follows the grass season, more strongly away from the equator.
• Lead times and release timesProduction lead time, quality release (test and hold), warehouse handling and transit between sites shift every requirement earlier. A lead-time ladder shows how far ahead each tier must act.
• From shortfall to decisionA heatmap of every site across 46 periods, the markets and customers behind any site's load, the capacity each site must add to stay under a target, and any two plans compared side by side.
• A network you drawThe Network Studio shows the chain by tier, with production sites, warehouses, consumption points and transport routes told apart. Add sites, draw routes between neighbouring tiers, drag sites into order, edit them in an inspector, define your own parameters and save the diagram as SVG or PNG; or work on a world map, with routes along their great circles. Every change re-plans at once and can be undone.
• Your own supply chainLoad a network from CSV, Excel or JSON (a template is one click away), edit every site, lane and setting, rename the tiers for another kind of chain, and share the whole plan as a link. Files stay in your browser.
• Planning within capacityA linear program, solved in your browser by HiGHS, plans the 13 years with no site above its capacity: it builds stock ahead of peaks, re-sources to suppliers with room and shows what reaches the shelves, market by market, and where the bottlenecks are.
• Learn it in five levelsA guided tutorial (beginner, intermediate, build a network, advanced, expert) with questions, in English, Spanish and Danish.

How it is built

The engine is a time-phased requirements plan, worked tier by tier from the consumer upstream over 156 months. A warehouse ships what its markets need and builds or draws safety stock. A CMO makes its output the release time early, grossed up for yield. Its inputs must arrive the lead time earlier still. Needs are split across suppliers by share, shifted by each lane's transit time and converted through the bill of materials. Yearly periods are built from their months, and each keeps its busiest month as well as its average. The plan is linear, so each site's load splits exactly into the markets it serves. Tests check the bill of materials chain, the time shifts, conservation of volume, safety stock, sourcing shares, the seasons of both hemispheres, expansions and outages, the exact tracing, the capacity to add, every file format and every translation. Within capacity, a linear program over the whole horizon decides what each site makes, which lane carries it and what stock is built ahead, at the least cost of unmet demand, and is checked on small chains whose answers are known. The capacity investment optimiser decides which candidate expansions are worth building, and about when, weighing discounted capex against the margin lost to unmet demand. Everything runs in your browser. The validation report re-runs every known-answer case live, from the same code.

Author

Tierwise was designed and built by Daniel Alberto Sepulveda Estay, PhD, founder of Estay Dynamics. He works on system dynamics and network optimisation for supply chain resilience. He holds a PhD from the Technical University of Denmark and an MSc from the Massachusetts Institute of Technology, and is a council member of the System Dynamics Society.

More about Daniel →

To cite: Sepulveda Estay, D. A. (2026). Tierwise: multi-tier capacity planning for a food supply chain (version 2.4.1) [Software]. https://kathuman.github.io/claude-projects/tierwise/