Initiatives · Digital Twins
Ontology-led models of assets, operations and dependencies, designed to compare feasible decisions as operating conditions change.
Current statusResearch initiative
Test the consequences of a decision against current evidence, capacity and timing before committing resources.
Digital Twins researches operational representations that link source records and observations to engineering, forecasting and planning models. An ontology gives assets, activities and dependencies consistent meaning; validated models describe how their state may change.
A maintenance delay, supply interruption or capacity change can affect commitments elsewhere. Separate dashboards and static models may miss those dependencies or compare options using inconsistent assumptions.
Define the system boundary and decision, map authorised inputs and establish a validated baseline. Specify how frequently state is updated, reconcile observations and expose stale or missing inputs. Compare scenarios against capacity, timing and policy constraints, keeping simulated changes separate from live records. A model becomes useful as a twin when its relationship to the operation is maintained and validated.
Government planners, infrastructure owners, industrial operators and supply-chain teams evaluating maintenance, allocation, disruption and investment decisions.
Connect assets, resources, processes and commitments through shared entities. Preserve record ownership, provenance, units and time references so connected models use compatible inputs.
Define update frequency and acceptable source age for the decision. Track conflicting observations, missing measurements and uncertainty rather than presenting every modelled state as a confirmed fact.
Compare outages, demand shifts and allocations against capacity, lead times, qualified substitutes and policy. Show trade-offs and infeasible options; the preferred option depends on stated objectives and model assumptions.
Compare predicted and observed outcomes, record model versions and investigate material discrepancies. Revalidation is required when operating conditions or model assumptions change; automatic model improvement is not assumed.
A delayed replacement component may postpone maintenance and reduce service capacity. A scoped twin would compare a qualified substitute, stock transfer and revised maintenance schedule, showing the resulting cost, timing and service consequences. Actual deliveries and maintenance outcomes would then test those estimates.
Assess input freshness, model error and uncertainty against representative operating data. Compare proposed decisions with an agreed baseline, including disruption cases and behaviour outside the validated operating range. Visual detail alone does not demonstrate predictive accuracy.
Begin with one system, a decision and measurable acceptance criteria. This initiative focuses on representation, simulation and optimisation; LTKAU develops investigation, readiness and governed response. Operational changes require separately defined integrations, permissions and approval.