| Who it is for | An operations or engineering leader who has a decision to make about a physical system and needs to know whether modelling it is the right way to make it. | An organisation that has completed Discovery and wants a working model of one process before committing to a wider deployment. | An operation ready to use a twin routinely rather than for one study: capacity planning, scheduling questions, change evaluation and investment cases. | An organisation running several operations that affect each other, where the interesting questions are the ones that cross a site or a system boundary. | Organisations whose requirements are set by their estate, their security posture and their obligations rather than by a scope of work. |
|---|
| Typical scope | One operation, or one clearly bounded part of a larger one. Two to four weeks depending on site access and data availability. | One production line, one area, one terminal process or one equivalent bounded operation. Typically six to twelve weeks. | A complete operation, or several connected ones. Typically three to six months, sequenced so parts of the model are usable before the whole is finished. | Multi-site or multi-system. Typically six months and beyond, delivered in phases with each phase independently useful. | Set in Discovery. There is no artificial ceiling. |
|---|
| Deployment model | No deployment. Discovery produces a document and a decision. | Single deployment on your infrastructure, or desktop delivery where that suits the pilot better. | Self-hosted or private cloud, inside your boundary. Access model agreed during Discovery. | Enterprise deployment inside your boundary, to your security and access requirements. | Whatever your obligations require, agreed in writing before work begins. |
|---|
| Simulation | Feasibility analysis only. Where it is useful to the assessment, a small illustrative model may be built to test a specific assumption. | Discrete-event and agent-based execution, seeded and replicated, with results reported as distributions. | Full discrete-event, agent-based and hybrid execution, with replication, seeding and versioned inputs. | Everything above, plus coupled multi-subsystem models and distributed execution of large run sets. | Scoped against the models, the run volumes and the compute the estate needs. |
|---|
| Data integration | Assessment of the systems that hold your operational data and what integrating them would involve. No integration is built. | Historical data import, and one live read path where the pilot requires it. | Live read paths from the systems of record for the modelled operation. The connectors required are scoped per site and the boundary is set in Discovery. | Multiple systems of record, industrial protocols and historical stores, scoped and built per site. | Scoped against the systems, protocols and boundaries actually in place. |
|---|
| Support | Included for the duration of the engagement. | Twelve-month warranty on the delivered work. Annual platform and support agreement optional from delivery. | Twelve-month warranty. Annual platform and support agreement, priced against the deployment. | Twelve-month warranty, an annual platform agreement, and a support agreement with agreed response commitments. | Agreed service levels, support tiers and escalation, written into the agreement rather than published as a claim. |
|---|
| Commercial model | Fixed price, agreed before work starts. Independent of anything that follows: the output is yours whether or not you proceed, and it is deliberately written so another supplier could act on it. | Fixed price against the scope agreed at the end of Discovery. Where a pilot proceeds to an operational deployment, the model built here is the foundation of it rather than being rebuilt. | Fixed price per phase against an agreed scope, with the annual agreement beginning at go-live. | Phased, against a programme scope set in Discovery. Optimization and decision-support components are scoped explicitly against their published maturity, and nothing at Research stage is sold as though it were finished. | By quotation after Discovery. Pricing depends on scope, number of assets and processes modelled, integrations, deployment model, simulation complexity, users, compute, support requirements, security requirements and service levels. |
|---|
