ONE CORE · MANY DOMAINS · CONTEXT-DEPENDENT OPTIMISATION

OPTIMIZE & IMPROVE

Factories, enterprises, logistics networks, energy systems, buildings, robotics and service organisations do not have the same processes — but they share the same challenge: turn reliable evidence into the right action at the right time.

FACTORYProduction · Quality · Energy · Maintenance
ENTERPRISESales · Finance · Documents · Human Work
LOGISTICSFleet · Routing · Warehouse · Delivery
ENERGYGeneration · Load · Cost · Carbon
ROBOTICSFleet · Tasks · Safety · Human Coordination
VALUE PLAN · MEASURED, NOT PROMISED

Benefits are planned against capacity, resources and process complexity.

Before a Analysed we define the available data, operational capacity, resource constraints and complexity of the target process. From this we build a measurable improvement plan with a baseline, target range, implementation effort and verification period.

CAPACITYAvailable machines, people, shifts, throughput headroom, energy limits.
RESOURCESEngineering time, data access, integration effort, maintenance and operating budget.
COMPLEXITYNumber of systems, dependencies, process variability, uncertainty and governance constraints.

Illustrative comparison plot

Indexed example only: current operation = 100. Lower is better for cost, energy, CO₂, errors and downtime. The target bars show an example planning envelope, not a guaranteed result.

Energy / unit
90–96
CO₂ / unit
91–97
Operational cost
92–97
Downtime
88–95
Internal errors
86–95
Baseline = 100Illustrative target envelope

Analysed planning scenarios

ScenarioTypical conditionPlanning effect
Low-accessFew sources, limited engineering timeObserve + identify opportunities
StandardUsable history, stable process, moderate integrationMeasure + rank + validate actions
AdvancedGood data coverage, cross-system access, active operator supportCross-domain orchestration + controlled execution

The exact KPI set is selected per plant: energy/unit, peak demand, CO₂/unit, OEE, downtime, scrap/rework, maintenance cost, throughput, labour hours, decision delay or other agreed indicators.

UNIVERSAL ARCHITECTURE

One optimisation core. Different domain models.

POLx°IMAT OOEP does not force every organisation into one factory-shaped or office-shaped model. A universal core handles evidence, state, uncertainty, optimisation, governance and execution while domain packs define the entities, processes, KPIs, constraints and actions that matter in each environment.

Incoming Data / EventCRM · ERP · MES · SCADA · sensors · documents · human input
Data Contract & Evidence GatewayIdentity · provenance · timestamp · schema · semantics · quality
Evidence ResolutionAccepted evidence · quarantine · conflict resolution · information requests
HIS Intelligence · Unified StateProcess context · dependencies · uncertainty · reliable horizon · state version
Local Domain DecisionsProduction · Energy · Quality · Maintenance · Sales · Finance · Documents
Cross-Domain DependenciesShared resources · constraints · deadlines · capacity · competing objectives
Candidate Action Set → Dynamic Priority + HMCDIOEEvaluate alternatives and trade-offs against current state and admissible constraints
HIS / EGU GovernanceEvidence sufficiency · robustness · authority · safety · Anti-AI-Cliff
Governed NEXT ACT / HCMOLRUN_NOW · PREPARE · SUGGEST · WAIT · REQUEST_EVIDENCE · HUMAN_REVIEW · ESCALATE · BLOCKED
Outcome → Verify / LearnMeasured result · decision audit · updated evidence · process learning
↺ New evidence updates the affected state, decisions and priorities. Independent branches continue.
DOMAIN PACKS

Optimisation adapts to the organisation.

The core remains stable. The domain pack changes what is observed, what is optimised, which constraints apply and which actions are available.

Industrial / Factory

Optimise interacting production, quality, energy, maintenance, material, labour and risk constraints.

ERP/MESSCADA/PLCOEEEnergyQuality
Orders → Machines → Production → Quality → Energy → Maintenance → Delivery

Enterprise / Office / Service

Optimise sales flow, decisions, document movement, finance, workloads, waiting time and lost opportunities.

CRMDMSFinanceProjectsPeople
Lead → Qualification → Decision → Documents → Finance → Outcome

Logistics & Warehouse

Balance routing, fleet utilisation, warehouse state, delivery time, energy, cost and service constraints.

WMSTMSFleetRouteCapacity
Demand → Inventory → Pick → Route → Fleet → Delivery → Verify

Energy & Utilities

Optimise generation, demand, storage, tariff exposure, carbon intensity, reliability and maintenance.

MeteringLoadStorageTariffsCO₂
Measure → Forecast → Balance → Dispatch → Verify

Buildings & Facilities

Coordinate HVAC, occupancy, comfort, maintenance, energy use, schedules and operating cost.

BMSHVACOccupancyComfortMaintenance
Occupancy → Demand → HVAC → Comfort → Energy → Maintenance

Robotics & Automation

Optimise task allocation, fleet state, exceptions, maintenance context and supervised human-machine execution.

RaaSFleetTasksSafetyHuman-in-loop
Task → Candidate Robot → Constraints → Allocation → Execute → Verify

Healthcare Operations

Operational optimisation can coordinate resources, scheduling, logistics and service flow while clinical decisions remain under appropriate professional authority.

SchedulingResourcesFlowEvidenceAuthority
Demand → Capacity → Schedule → Service Flow → Outcome

Construction & Projects

Coordinate schedules, subcontractors, materials, cost, milestones, documents, risk and site constraints.

ProjectCostMaterialScheduleRisk
Plan → Resource → Execute → Inspect → Re-plan → Deliver

Custom Domain

Where a process has measurable state, constraints, actions and outcomes, a new domain pack can be mapped onto the same OOEP core.

OntologyCustom KPIRulesConnectorsActions
Discover → Model → Connect → Optimise → Govern → Verify
INDUSTRIAL VALUE · PARTNER DISCUSSION

Connect the plant. Improve the whole operation.

A proposed industrial Analysed for an engineering partner such as Bajorath: connect available production and energy evidence, understand cross-system dependencies, and evaluate improvements before changing operations. The diagram shows the intended architecture, not an existing customer installation.

PLANT & BUSINESS SYSTEMSMachines / PLC / SCADA
MES / ERP / Orders
Energy meters / tariffs
Quality / maintenance
Warehouse / people
POLx°IMAT OOEPCONNECT → EVIDENCE → UNIFIED STATE
HIS Intelligence → HMCDIOE
HIS/EGU → NEXT ACT → HCMOL
VERIFY / LEARN
GOVERNED IMPROVEMENTSOperator recommendations
Production / energy scheduling
Maintenance prioritisation
Quality / exception routing
Human-approved actions
LESS ENERGYReduce avoidable consumption and improve energy use per unit of output.
LESS CO₂Reduce emissions where measured energy savings and the applicable emission factors support the result.
LOWER COSTReduce energy, downtime, waste and avoidable operational expenditure.
BETTER PERFORMANCEImprove throughput, quality, reliability and resource utilisation within safe constraints.
Energy efficiency

Identify idle consumption, inefficient operating windows, peak-load exposure and opportunities to coordinate production with energy demand.

CO₂ reduction

Connect energy and process evidence to emissions accounting and evaluate lower-carbon operating alternatives without inventing savings.

Budget & operating cost

Compare energy, labour, maintenance, material and delay costs so local savings do not create larger costs elsewhere.

Fewer internal errors

Detect missing, conflicting or outdated information; improve handovers, exception routing and verification of completed actions.

Less downtime & waste

Prioritise maintenance and process interventions using available evidence, dependencies and expected operational impact.

Better decisions

Give engineers and operators a traceable NEXT ACT with evidence, uncertainty, constraints, responsible owner and expected outcome.

Analysed measurement: establish a baseline first, then compare energy per unit, CO₂, operating cost, downtime, scrap/rework, throughput and decision/response time where relevant data exists. No percentage improvement is claimed before validation. Initial operation is read-only; any physical control requires separate safety, authority and integration approval.
CORE MODULES

Build only the optimisation stack you need.

Start with evidence and observation, then add process intelligence, optimisation, orchestration and controlled execution as the customer environment becomes validated.

01 · CONNECT

Register systems, APIs, files, machines, sensors and external sources.

source → contract
02 · EVIDENCE

Preserve source, identity, time, provenance, quality and raw references.

fact ≠ inference
03 · SEMANTICS

Normalize schema, units, IDs and business/process meaning across systems.

raw → ontology
04 · UNIFIED STATE

Create current cross-system state for cases, processes, resources and risks.

E(t) → State(t)
05 · HIS INTELLIGENCE

Understand process state, dependencies, deviations, unknowns and opportunity.

know → interpret
06 · PRIORITY ENGINE

Re-rank what matters now as deadlines, capacity, risk and events change.

Priorityᵢ(t)
07 · HMCDIOE

Evaluate interacting objectives and candidate actions under constraints.

argmax A_adm
08 · HIS / EGU

Govern evidence sufficiency, uncertainty, reliable horizon and authority.

admissibility gate
09 · HCMOL

Route tasks to people, software or permitted machine actions with approval.

NEXT ACT → execute
10 · ORCHESTRATION

Coordinate dependencies across departments, sites, machines and vendors.

cross-domain runtime
11 · VERIFY / LEARN

Compare predicted and realised outcomes and update the next cycle.

before ↔ after
12 · OPERATOR CONTROL

Human supervision for runtime, questions, incidents, approvals and audit.

observe · mentor · govern
DOMAIN MATRIX

Same core — different optimisation objective.

DomainPrimary StateTypical ObjectiveKey ConstraintsExecution Target
FactoryMachines, orders, quality, energy, maintenanceThroughput + quality − energy − downtime − waste − riskSafety, process envelope, capacity, qualityEngineer, MES, operator, permitted OT
EnterpriseCustomers, deals, people, documents, financeRevenue + conversion + productivity − waiting − cost − riskAuthority, compliance, capacity, evidenceEmployee, CRM, DMS, finance workflow
LogisticsFleet, routes, inventory, orders, capacityService + utilisation − delay − distance − cost − riskDelivery windows, capacity, traffic, regulationDispatcher, TMS, WMS, fleet
EnergyGeneration, load, storage, tariff, carbonReliability + value − cost − carbon − imbalanceGrid limits, equipment, reserve, tariffEnergy operator, scheduler, controller
BuildingsOccupancy, HVAC, comfort, energy, faultsComfort + availability − energy − cost − failure riskComfort bands, schedules, equipment limitsFacility team, BMS, maintenance
RoboticsRobot state, tasks, battery, location, exceptionsTask completion + utilisation − idle − risk − energySafety, workspace, battery, permissionsFleet manager, robot, human supervisor
RUNTIME MODEL

Optimisation is continuous, not a one-time recommendation.

Every relevant event can update the state, change dependencies, alter priorities and create a new admissible NEXT ACT set. Independent branches can continue while blocked branches wait for evidence or approval.

EVENT
EVIDENCE
STATE
HIS INTELLIGENCE
PRIORITY
HMCDIOE
HIS / EGU
HCMOL
VERIFY
The objective is not maximum automation. The objective is maximum useful, evidence-supported improvement with controlled risk. When information is insufficient, the correct runtime action may be REQUEST EVIDENCE, WAIT, HUMAN REVIEW or ESCALATE.
DEPLOYMENT PATH

Start small. Validate. Expand.

Analysed / Reference Case

  • One site, process, line or enterprise workflow
  • Read-only evidence collection
  • Baseline and process discovery
  • Shadow recommendations
  • Human-approved NEXT ACT
  • Measured before / after verification

Scale / Enterprise Runtime

  • Multiple departments, sites and systems
  • Shared control plane + tenant isolation
  • Cross-domain priorities and dependencies
  • Operator control and incident handling
  • Reusable client / industry configuration packs
  • Controlled expansion of execution authority