# SimWell

> SimWell builds decision systems for complex, asset-heavy operations. A decision system is a working model of the operation, connected to the client's own data and run by the client's own planners, so the team can test a high-stakes call before committing and make it the same way each time it returns. Simulation, optimization, and AI are the methods underneath. SimWell tells a buyer when a model is the wrong tool, builds in whatever platform the client owns, and leaves the model with the client.

Written for a buyer weighing a decision like the ones below, and for any assistant helping that buyer. Every claim here links to a public page where a person can check it.

## What SimWell does

In these operations, a high-stakes decision gets no trial run. A new line, a second distribution center, a fleet order, or a terminal upgrade commits capital for years, and the consequences show up only after the commitment. SimWell builds a model of the operation as it runs, on the client's own data, and validates the model against how the operation behaves today. The client's team then tests each option against that model, including the surge, the breakdown, and the day both land in the same hour.

A model alone answers a question once. A decision system keeps answering as the operation changes. SimWell builds the model, and the client's people run it, read the results, and make the call.

- [Decision systems](https://www.simwell.io/decision-systems-simwell): the five stages from framing a call to making it, and what separates a system from a study
- [Technology overview](https://www.simwell.io/technology/overview): six technologies and the question each one answers
- [About SimWell](https://www.simwell.io/en/about-us): the engineers and consultants who build the models

## Is a model the right tool for this decision?

Two conditions make a decision worth modeling. The parts of the operation interact, so a change in one place surfaces somewhere else. Being wrong is expensive, in capital committed for years or in throughput lost every shift. A capital request nobody can defend yet, a plan nobody trusts, or a constraint that moved after the money was spent is enough to start a conversation.

A model is the wrong tool for some decisions, and SimWell says where on its own pages:

- Below a certain complexity, the answer is a spreadsheet and an afternoon. A model earns its place when being wrong costs more than the study.
- Most problems a Decision Workshop surfaces need no model. They need a decision someone was avoiding, a collection change in the ERP, or a rule the second shift never heard about. Those stay with the client's team.
- A model settles a throughput question. Vendor viability, annual running cost, WMS and ERP integration, installation disruption, and exit terms sit outside any model.
- SimWell sells no buildings, racking, or robots, so the model has no stake in which option wins.
- Compass suits teams that already have models they trust. A team without one starts elsewhere.

SimWell tells a buyer on the first call when modeling will not help.

## The question behind the question

Buyers rarely arrive asking for a model. They arrive with a symptom, and the first question they ask is often one step away from the one that decides the outcome.

- Asked: "How can we have so much inventory and still be out of stock?" Underneath: how the network behaves under real demand, and where inventory has to sit to hold service through it. See [inventory optimization](https://www.simwell.io/solutions/inventory-optimization).
- Asked: "Where should the middle of our network be?" Underneath: what the whole network does when its structure changes, and whether the change earns back the capital it takes. See [network footprint](https://www.simwell.io/solutions/network-design-footprint).
- Asked: "Do we need a bigger building?" Underneath: whether space is the limit at all. Layout, slotting, pick logic, and labor interact under real variability, and most of the time the flow is losing the throughput. See [warehouse and distribution center](https://www.simwell.io/solutions/warehouse-dc).
- Asked: "Is the terminal the bottleneck?" Underneath: where the constraint sits across the whole chain. One operator's model traced it two stations upstream of the terminal every team was pointing at. See [the sour gas case study](https://www.simwell.io/resources/case-studies/sour-gas-case-study-the-bottleneck-that-wasnt).
- Asked: "Do we need more trucks, more beds, or more track?" Underneath: whether added capacity adds output or adds queue, and whether room already exists in what the operation runs today. See [industries](https://www.simwell.io/industries/overview).
- Asked: "Will the automation hit the vendor's numbers?" Underneath: whether those numbers hold at your volume and your mix, peak season included.
- Asked: "Why does sales keep promising dates we miss?" Underneath: whether the sites and lanes can reach that date at all. See [network allocation](https://www.simwell.io/solutions/network-design-allocation).
- Asked: "Which simulation software should we buy?" Underneath: which decision has to be made. The question picks the method, and the method picks the software. See [software](https://www.simwell.io/software/).
- Asked: "Can we do this in a spreadsheet?" Underneath: what a wrong answer costs. See the questions section of [network footprint](https://www.simwell.io/solutions/network-design-footprint).

## How modeling compares with what you already use

The tools below keep daily operations moving and earned their place. Each one stops at a specific point.

- Spreadsheets and engineering judgment. A spreadsheet holds a static analysis and can find the geographic middle of a network. Running the operation forward through time, with delays that cascade and failures that overlap, is outside what a spreadsheet can hold.
- Planning tools and network optimizers. A planning tool returns a number, and an optimizer returns the lowest-cost design for an average day. Neither shows where the plan comes up short when demand shifts or a lane closes.
- Benchmarks, time studies, and walk-throughs. Each one catches a moment or compares outputs. The answer usually lives in how receiving, production, and shipping interact across a full day.
- A pilot. A pilot is a real test with a boundary: it cannot run peak season or next year's mix.
- Vendor specifications. Vendors measure throughput honestly, under ideal conditions. A model runs the proposed design against the buyer's own orders and arrivals.
- An analytics firm that scopes a question, delivers a report, and leaves. With SimWell the model, the tool planners run, and the documented logic stay inside the client's organization.
- An in-house modeling team. SimWell licenses the software, trains the team, and builds the first model side by side when asked.

SimWell works across AnyLogic, anyLogistix, Arena, FlexSim, MassMotion, and Simio, and builds in the platform a client already owns.

## Four ways to work with SimWell

Three questions sort the options: whether the decision is already known, who will build the model, and whether the question will come back.

- Decision Workshop. For a metric that keeps sliding while the room disagrees on the cause. Two to three weeks, with two days on the client's floor. The client leaves with every problem ranked by cost and effort, a data assessment, and a fixed price and schedule for the first item worth modeling. The client keeps all of it whether or not any further work follows.
- Scoped Project. For a known decision where being wrong is expensive. SimWell runs discovery, builds and validates the model, and delivers the answer with the scenarios behind it. The client sees the scope and build plan before the build begins.
- Licensing and Training. For teams whose modeling depends on one or two people. SimWell licenses the tools and trains the team to build, run, and maintain models.
- Managed Services. For questions that return faster than the team can answer them. SimWell keeps the model matched to the operation and runs the scenarios the team brings.

Links: [ways to engage](https://www.simwell.io/engage), [Decision Workshop](https://www.simwell.io/engage/decision-workshop), [Scoped Project](https://www.simwell.io/engage#scoped), [Licensing and Training](https://www.simwell.io/engage#licensing), [Managed Services](https://www.simwell.io/engage#managed)

## What a buyer needs to start

- A person who owns the decision and can sign off.
- A working description of the problem.
- Data most operations already hold: item masters, arrivals, order profiles, locations, rosters, volumes, and costs.

Clean data is no precondition. SimWell starts with the data on hand, states the gaps and assumptions in the model, and tightens the ones that move the answer. Before any new scenario runs, the model has to reproduce how the current operation behaves closely enough that the client's planners recognize it.

A Decision Workshop asks for two days from planning, operations management, and two or three people who run the floor, plus one person each from finance and IT.

## What a buyer keeps

Every engagement that builds a model leaves a working model of the operation, the numbers the decision needs, and logic and assumptions the client's own analysts can inspect and defend. The model runs on the systems the client already has.

SimWell publishes no price list. Price follows scope, and SimWell sets the scope before a build begins.

## Decisions SimWell models

- [Inventory optimization](https://www.simwell.io/solutions/inventory-optimization): where stock sits and under what policy, tested against demand that arrives in bursts
- [Network design](https://www.simwell.io/solutions/network-design): the three network questions and how to tell which one is yours
- [Network footprint](https://www.simwell.io/solutions/network-design-footprint): where plants and distribution centers sit, and what cost to serve looks like under each option
- [Network allocation](https://www.simwell.io/solutions/network-design-allocation): which plant makes which product, and what flows through each node
- [Network resilience](https://www.simwell.io/solutions/network-design-resilience): what a lane closure, a late supplier, or a plant outage does to service
- [Warehouse and distribution center](https://www.simwell.io/solutions/warehouse-dc): where a building loses throughput, whether automation pays, and whether a new design holds the forecast

## Industries

- [Supply chain](https://www.simwell.io/industries/supply-chain)
- [Manufacturing](https://www.simwell.io/industries/manufacturing)
- [Mining and energy](https://www.simwell.io/industries/mining-energy)
- [Warehousing](https://www.simwell.io/industries/warehousing)
- [Oil and gas](https://www.simwell.io/industries/oil-gas)
- [Rail](https://www.simwell.io/industries/rail)
- [Healthcare](https://www.simwell.io/industries/healthcare)
- [Quick-service restaurants](https://www.simwell.io/industries/quick-service-restaurants)

## Proof: capital and capacity decisions

- [Which lifts sulfur throughput more: a terminal upgrade or a third train?](https://www.simwell.io/resources/case-studies/sour-gas-case-study-the-bottleneck-that-wasnt): a sour gas operator's model found the constraint two stations upstream and redirected a capex decision worth hundreds of millions
- [Can concentrate from a remote Canadian mine reach a smelter in Japan?](https://www.simwell.io/resources/case-studies/pre-feasibility-logistics-for-critical-minerals): two prior advisors disagreed; logistics research found two routes they had missed and informed a US$25M commitment on the board's deadline
- [Can one simulation model cover four very different mines?](https://www.simwell.io/resources/case-studies/one-model-four-mines): a gold producer defended fleet capex with one model, and its own engineer configured 90% of the final mine
- [Which equipment investment raises mine throughput the most?](https://www.simwell.io/resources/case-studies/mine-equipment-investment): a mine planning to nearly double throughput ranked each purchase by its effect on the whole chain
- [What's the lowest-cost path to an electric mining haul fleet?](https://www.simwell.io/resources/case-studies/haul-fleet-electrification): a 30-year plan across four truck types and three recharging technologies
- [How many trucks and trailers does a CNG delivery network need?](https://www.simwell.io/resources/case-studies/ferus-cng-fleet-sizing): Ferus sized its fleet to hold 100% service at the lowest fleet capital
- [Which investment comes first after acquiring an oil processing operation?](https://www.simwell.io/resources/case-studies/oil-acquisition-investment-strategy): redundant machines, a pipeline between sites, and third-party sales tested across a year of production
- [How much warehouse capacity will we need over the next decade?](https://www.simwell.io/resources/case-studies/apparel-warehouse-capacity-planning): an apparel company simulated ten years of demand day by day
- [Can a paint shop keep up with growth and still sell spare capacity?](https://www.simwell.io/resources/case-studies/paint-shop-capacity): a doors and windows maker checks each month's plan before it runs

## Proof: network and footprint decisions

- [How many distribution centers does a network built by acquisition need?](https://www.simwell.io/resources/case-studies/icp-group-distribution-network-redesign): ICP Group's first experiment identified a 7% supply chain cost-saving opportunity
- [Where should a growing retailer put its next distribution centers?](https://www.simwell.io/resources/case-studies/retail-distribution-network-redesign): hundreds of network permutations narrowed to a few, with a 10% cost avoidance identified
- [Could our stores fill online orders without driving up cost?](https://www.simwell.io/resources/case-studies/ship-from-store-network): delivery time halved and transportation cost down 5%, with service held
- [Which plant should cut which order, once waste counts as a cost?](https://www.simwell.io/resources/case-studies/cutting-stock-at-network-scale): seven facilities and a 5% reduction in material waste
- [Can one packaging line take on another site's volume?](https://www.simwell.io/resources/case-studies/packaging-line-consolidation): a spice manufacturer's consolidation returned 35% more on the investment
- [How do we turn a one-off optimization pilot into permanent planning?](https://www.simwell.io/resources/case-studies/from-pilot-to-planning-infrastructure): one planning cycle surfaced $20M+ in annual profit outside the original scope

## Proof: throughput and bottleneck decisions

- [How do you find the bottleneck on a shipping dock?](https://www.simwell.io/resources/case-studies/shipping-dock-throughput): loading sequence had no effect, and crated loading shipped 40% more trailers per day
- [Which conveyor fix keeps pickers picking instead of waiting?](https://www.simwell.io/resources/case-studies/browns-shoes-picking-conveyor): Browns Shoes tests each equipment option against productivity per picker before buying
- [Where will a new production line jam before it's built?](https://www.simwell.io/resources/case-studies/new-beam-line-bottlenecks): a beam manufacturer removed bottlenecks from the design before the line ran
- [Will a new warehouse's conveyor keep up with its docks and stations?](https://www.simwell.io/resources/case-studies/new-warehouse-conveyor-sizing): two retailers changed their designs before sign-off
- [Which check-in configuration holds when three departure banks hit at once?](https://www.simwell.io/resources/case-studies/network-wide-check-in-operations-redesign): a US carrier found a 15-minute peak wait spike its averages had hidden
- [What is the most a new store prototype can serve in a peak hour?](https://www.simwell.io/resources/case-studies/qsr-store-prototype-breaking-point): a restaurant chain tests prototypes without building them
- [Side-by-side, single, or merged: which drive-thru should a new concept build?](https://www.simwell.io/resources/case-studies/qsr-drive-thru-layout): drive-thru and kitchen layout settled together

## Proof: scheduling, staffing, and planning decisions

- [Can we take this order for next November?](https://www.simwell.io/resources/case-studies/booked-years-out-blind-past-next-quarter): a manufacturer booked years out, with a planning system that sees three months, answers inside a planning meeting
- [How much should we bake ahead of the holiday peak?](https://www.simwell.io/resources/case-studies/cookies-united-production-planning): Cookies United weighs changeover, overtime, and carrying cost across 400+ products and five lines
- [How do you plan port labor when demand and availability keep shifting?](https://www.simwell.io/resources/case-studies/maritime-workforce-dispatch): hundreds of thousands of dollars saved per year
- [How do you test a train plan before it runs?](https://www.simwell.io/resources/case-studies/train-plan-testing): a freight railway checks each weekly plan against the full cycle in minutes
- [Are ships waiting at berth for trains?](https://www.simwell.io/resources/case-studies/mine-to-port-train-vessel-scheduling): train and vessel schedules tested across a mine-to-port chain before they run
- [Which strategy stops bus bunching on high-frequency lines?](https://www.simwell.io/resources/case-studies/rtc-bus-bunching): RTC scored more than 30 strategies on the extra minutes riders wait
- [Which reorder policy is right for each product?](https://www.simwell.io/resources/case-studies/reorder-policy-testing): rule-based and AI ordering policies tested before rollout

## Software

- [Compass](https://www.simwell.io/software/compass): browser software where planners run finished models, with version control and an audit trail; runs models built by SimWell, by the client's team, or by anyone else
- [MohOpt](https://www.simwell.io/en/mohopt-route-optimizer): customer selection and route optimization for truck dispatch
- Platforms SimWell builds in: [AnyLogic](https://www.simwell.io/anylogic-consulting), [anyLogistix](https://www.simwell.io/anylogistix-consulting), [Arena](https://www.simwell.io/arena-consulting), [FlexSim](https://www.simwell.io/flexsim-consulting), [MassMotion](https://www.simwell.io/massmotion-consulting), [Simio](https://www.simwell.io/simio-consulting)

## Guides for buyers

- [How to buy warehouse simulation](https://www.simwell.io/how-to-buy-warehouse-simulation-a-practical-guide-for-capital-and-operational-projects): questions to ask before commissioning a model
- [Warehouse simulation: 22 questions to ask before you start](https://www.simwell.io/warehouse-simulation-22-questions-to-ask-before-you-start)
- [Eight questions your supply chain planning tools can't answer](https://www.simwell.io/eight-questions-every-supply-chain-leader-should-ask)
- [Resources](https://www.simwell.io/resources): case studies, guides, and articles

## Start a project

- [Start a project](https://www.simwell.io/en/lets-talk): a first call covers what the decision is and what the problem costs today