Captia Technology

Glossary

Verifiable definitions for the industrial, AI and energy terms Captia Technology uses.

AI visual inspection: Use of industrial cameras and deep-learning models to detect defects, missing parts or quality deviations in real time. Captia AI deploys it at the edge to meet line latencies (<100 ms) and per-unit traceability.
Anomaly detection: Technique that identifies unusual behaviour in time series or industrial patterns without requiring prior labels. Captia AI combines rules, statistical models and deep learning depending on the customer’s data maturity.
API-first: Design approach where every feature is exposed first as a stable, versioned and documented API before any UI. Captia Service requires it in new integrations to avoid "glue-script" technical debt.
Battery energy storage (BESS): Battery Energy Storage System, the industrial batteries that absorb and release energy for arbitrage, backup or renewable integration. Captia Energy sizes them against the real plant profile, not the trend.
CANopen: CANopen is the application layer and set of device profiles that the CAN in Automation (CiA) association defines on top of the CAN bus, standardised as EN 50325-4. Every node publishes its functionality in an object dictionary addressed by index and sub-index, and exchanges process data in PDO frames of up to eight bytes.
Closed loop: A closed loop is a system in which the output is measured and used to correct the input without waiting for a person. In a digitalised plant it means the software does not just display the data: it decides and executes under supervision through rules, workflows, energy setpoints or maintenance work orders. It is level 3, Acting, of the Physical AI Ladder, and it requires connected, reliable data.
Corporate PPA: Power Purchase Agreement, a long-term contract between a company and a renewable producer that secures price and origin of the electricity consumed. Captia Energy considers it when self-consumption cannot cover the target footprint.
Data acquisition (DAQ): Continuous capture of signals from machines, sensors, meters and controllers and their conversion into digital data with a name, unit and timestamp. On the shop floor it requires multi-protocol reading (Modbus, OPC UA, MQTT), normalisation at the source and local buffering so no data is lost without connectivity. It is the input layer of Captia Connect.
Data lakehouse: Architecture that combines cheap data-lake storage with the ACID guarantees and analytic performance of a data warehouse (Delta, Iceberg, Hudi). Captia Connect uses it as the natural sink for normalised OT data.
Data lineage: Full traceability of a data point from its physical source (sensor) to the decision point (dashboard, model). Captia Connect keeps it so auditing or debugging a metric never requires archaeology.
Data observability: Ability to detect anomalies in freshness, volume, schema and distribution of data streams before they break a downstream decision. Captia Connect embeds data-quality tests and data SLOs as part of the rollout.
Decision horizon: The decision horizon is the time scale on which a decision must be taken and takes effect: from the regulatory loop in milliseconds to supervision in seconds, shift optimisation in minutes and planning in days. It orders where compute lives, because every decision should run at the highest level that meets its latency without preventing the level below from degrading safely.
Demand response (DR): Programme that rewards consumers for reducing or shifting consumption in critical grid windows. Captia Energy integrates DR into plant operation without compromising quality or production deadlines.
Digital twin: Live representation of a plant asset, process or system fed by real data and simulation. Captia integrates digital twins only when they drive a decision, never as a 3D showcase.
Edge AI: Edge AI is the execution of artificial intelligence models inside the plant, next to the machines, instead of in the cloud. It builds on edge computing: the same device that captures and normalises the data can run inference on it with low latency and without depending on connectivity. It is the path by which physical AI reaches sensors, cameras and controllers without moving sensitive data outside the perimeter.
Edge computing: Data processing close to its physical source, on the shop floor, instead of round-tripping to the cloud. Lowers latency, enables real-time decisions, and keeps sensitive data inside the perimeter.
Energy efficiency: Ability to produce the same output, or more, while consuming less energy. In industry it is measured by comparing energy consumed with output obtained (specific consumption) and improved by acting on equipment, processes, schedules and operating habits, with savings verified against a baseline.
Energy optimisation: Joint adjustment of setpoints, schedules and resources to minimise cost or energy footprint without harming production. Captia Energy combines predictive models with the customer’s real operating rules.
ERP: Enterprise Resource Planning, the integrated suite that manages finance, procurement, sales, inventory, manufacturing and HR. Captia Service deploys it as the backbone where operational decisions materialise.
EtherCAT: EtherCAT (Ethernet for Control Automation Technology) is an Ethernet-based fieldbus maintained by the EtherCAT Technology Group and specified in IEC 61158. Its defining trait is on-the-fly processing: one frame travels through every daisy-chained slave and each one reads and writes its own data as the frame passes, instead of storing and forwarding it like a switch.
EtherNet/IP: EtherNet/IP is the industrial network defined by ODVA that runs the CIP application protocol (Common Industrial Protocol) over standard Ethernet and TCP/IP. The IP stands for Industrial Protocol, not Internet Protocol. It combines on-demand explicit messaging over TCP with cyclic implicit messaging over UDP, whose rate is set by each connection’s RPI parameter.
EUROMAP 63: EUROMAP 63 is the data exchange interface between plastics injection moulding machines and supervisory systems, defined by the European association of plastics and rubber machinery manufacturers. It works through text files dropped in a shared network directory: the requesting system writes a job file and the machine answers with a response file.
EUROMAP 77: EUROMAP 77 is the OPC UA information model that standardises data exchange between plastics injection moulding machines and MES systems. It replaces the file-based exchange of EUROMAP 63 with an OPC UA server on the machine itself, offering defined objects, types and units, and subscription-based reads on value change.
Field service: Management of technical interventions on site: planning, mobile execution, spare parts and reporting. Captia Service digitalises it with Odoo flows plus industrial terminals and feeds it with AI alerts.
FINS: FINS (Factory Interface Network Service) is Omron’s proprietary protocol for reading and writing memory areas in its controllers. It runs over Ethernet in UDP and TCP variants, and also over serial links and the vendor’s own networks. Addressing combines network, node and unit numbers with the target memory area and word.
HART: HART (Highway Addressable Remote Transducer) is the protocol that superimposes a frequency-modulated digital signal on the 4-20 mA analogue loop of process instrumentation. The modulation has zero mean value, so the analogue measurement stays valid while digital traffic flows over the same pair. It is maintained by FieldComm Group.
Human-in-the-loop: Pattern that embeds explicit human validation in critical automated decisions: quality, process safety, worker safety. Captia AI designs it so operators gain context, not to replace them.
IEC 62443: Family of international standards for security in industrial automation and control systems (IACS). Captia Consulting uses them as the reference framework when a customer needs to certify OT security maturity.
Industrial AI agent: An industrial AI agent is an autonomous language-model-based system that observes the plant operational context, reasons over it and executes bounded actions under human supervision, such as opening a maintenance work order or reconfiguring a workflow. It differs from a chatbot in that it does not just answer: it accesses real sources (historian, MES, CMMS) and acts with explicit limits, defined permissions and full traceability.
Industrial automation: Industrial automation is the use of control systems (PLCs, robots, instrumentation) to run production processes with minimal manual intervention. It splits into process automation (continuous flows: chemicals, food) and discrete automation (part-based manufacturing), and is organised in layers: field control, supervision (SCADA) and data exploitation. That last layer is where Captia works.
Industrial change management: Discipline that coordinates communication, training and support so operators, middle management and leadership adopt new processes and technology. Captia Consulting embeds it in every roadmap wave.
Industrial computer vision: Application of computer vision on the shop floor: quality inspection, defect detection, counting and traceability. Captia AI deploys edge models on industrial cameras integrated by Connect.
Industrial data contracts: Explicit agreement on schema, SLA, quality and owner of a data stream between plant and IT consumers. Captia Connect formalises them with semantic versioning so no OT change breaks downstream analytics silently.
Industrial decarbonisation: Structured reduction of direct and indirect emissions (Scopes 1-3) of an industrial operation. Captia Energy addresses it by combining efficiency, electrification, self-consumption and green contracts backed by real M&V.
Industrial forecasting: Quantitative prediction of demand, energy consumption, supply or process parameters via statistical or deep-learning models. Captia AI combines it with real operational constraints (shifts, contracts, minimum stock).
Industrial gateway: An industrial gateway is the device that bridges the plant network and the information systems: it reads controllers, sensors and meters over their native protocols and republishes the data northbound in a single format. It concentrates protocol translation, timestamping and temporary buffering for when the link drops.
Industrial historian: Time-series database optimised for plant telemetry (OSIsoft PI, Aveva, InfluxDB, TimescaleDB). Captia Connect feeds it or replaces it with streaming architectures when historical cost is not justified.
Industrial IoT (IIoT): Network of sensors, actuators and devices specifically designed for industrial environments: ruggedness, latency and security beyond consumer IoT. Captia Connect covers its end-to-end integration.
Industrial proof of concept: Time-boxed, measurable experiment that validates a technological or operational hypothesis in real plant conditions. At Captia Consulting every PoC is designed with pre-agreed scaling criteria, so there are no "eternal PoCs".
Industrial self-consumption: Energy model where a company generates part of its electricity (typically solar) to cover its own consumption, reducing market exposure. Captia Energy designs and integrates the installation with operational systems.
Industrial SRE: Adaptation of Site Reliability Engineering to industrial contexts: SLOs grounded in production impact, runbooks that engage OT and blameless postmortems. Captia Service rolls it out together with the customer’s internal team.
Industrial systems integrator: An industrial systems integrator is the company that connects a plant’s equipment and software so they work as one system: PLCs, SCADA, MES, ERP and the data layer. Unlike a consulting engineering firm, which diagnoses and recommends, the integrator implements, configures and leaves the system running in production. Captia combines both roles: diagnosis first, integration afterwards.
Industrial timestamp: An industrial timestamp is the instant attached to every piece of plant data, set as close to the sensor as possible and never at ingestion time. It is what fixes the causal order of events across different machines and what makes that order survive network delays. Without a common clock and a reliable timestamp, the history lets you count, not explain.
Industry 4.0: Industrial paradigm built on data integration, IoT, AI and automation to turn the plant into a cyber-physical system connected to the business. Captia implements it as one system, not a collection of pilots.
Industry 5.0: Industry 5.0 is the framework promoted by the European Commission that adds three priorities on top of Industry 4.0: putting people at the centre (technology assists the operator rather than replacing them), resilience of production chains, and sustainability. It does not replace 4.0: it redirects its technologies towards those goals instead of efficiency as the sole objective.
IO-Link: IO-Link is the point-to-point digital communication interface between a master and a sensor or actuator, standardised as IEC 61131-9. It replaces the analogue or binary signal of the last stretch of cable with a three-wire serial link that carries configuration parameters, device identity and diagnostics on top of the process value.
IPMVP (energy M&V): International Performance Measurement and Verification Protocol, the standard framework to prove real energy savings against a baseline. Captia Energy applies it to every significant improvement to remove contractual ambiguity.
ISO 50001: International standard to implement an energy-management system (EnMS) with measurable continuous improvement. Captia Energy uses it as the organisational scaffolding so energy improvements outlive shift and leadership changes.
MES: Manufacturing Execution System. It orchestrates shop-floor execution between control (PLC/SCADA) and planning (ERP). Captia integrates MES as the middle layer, fed by Connect and enriched by AI.
MLOps: Practices that take machine-learning models to production with traceability, monitoring and managed lifecycle. Captia AI adapts it to industrial environments: few labels, sensitive data, frequent drift.
Modbus: Modbus is the most widespread industrial communication protocol: published by Modicon in 1979 and royalty-free, it remains the de facto interface of drives, meters, power analysers and controllers. A client reads and writes numbered registers and coils on a server, with no names, types or encryption: what each address means comes from the vendor’s documentation.
Model drift: Degradation of a production model when the data distribution shifts from training. Captia AI monitors drift with statistical tests and operational alarms, not only ML metrics.
MQTT: Lightweight pub/sub protocol designed for IoT and reliable communication over constrained bandwidth. Captia Connect uses it to transport sensor telemetry to the edge and cloud.
MRP II: Manufacturing Resource Planning, the integrated planning of materials, capacity and production inside the ERP. Captia Service configures it in Odoo against the plant’s real routing rules and bottlenecks.
MTConnect: MTConnect is an open, royalty-free standard for publishing machine-tool data as XML over HTTP, published as an ANSI standard and maintained by the MTConnect Institute. It is read-only by design: an adapter on the machine feeds an agent that exposes the data with standardised names, types and units, with no ability to command the machine.
NIS2 directive: European directive (EU 2022/2555) that extends cybersecurity obligations to essential and important sectors, including manufacturing and energy. Captia Consulting treats it as a design constraint in any OT/IT project.
Odoo: Open-source ERP suite used by Captia Service as the base platform to digitalise finance, inventory, CRM, manufacturing and business workflows. Captia implements, customises and maintains it with its own engineering team.
OEE: Overall Equipment Effectiveness, a metric that combines the availability, performance and quality of a machine or line. Captia Consulting uses it as the starting point to prioritise measurable improvements before proposing technology.
Official Odoo partner: An official Odoo partner is a company recognised by the vendor to work with its ERP, with access to its material, its support and its criteria. That relationship gives a direct channel with the vendor; on its own it does not determine the outcome of an implementation, which depends on prior process analysis. Captia Technology is an official Odoo partner: it implements, customises and maintains the ERP with its own engineering team, and trains the teams that use it.
OPC DA: OPC DA (Data Access) is the classic OPC specification that defines real-time access to process values through Microsoft COM/DCOM technology. Each data point is identified by an ItemID string and returned with value, quality and timestamp. The OPC Foundation treats it as legacy and directs new development to OPC UA.
OPC-UA: Vendor-neutral industrial communication standard (IEC 62541) for exchanging data between OT equipment and IT systems. Captia Connect uses it as the preferred protocol for normalised ingestion.
Operational dashboard: Screen designed for in-shift decisions, not for monthly meetings. Captia AI builds them tied to the specific KPI and operator who acts on it: fewer charts, more context.
Operational fragmentation: State where a company's machines, sensors, energy, ERP, processes and operations run disconnected, producing partial data, slow decisions and transformations that never complete. It is the problem Captia Technology solves by operating them as one system.
Operational generative AI: Applications of foundation models (LLMs) to real industrial business processes: assisted reporting, internal search over technical documentation, operator copilots. Captia Service rolls them out with guardrails and active auditing.
Operational KPI: Quantitative indicator that measures a critical attribute of an operation, such as OEE, throughput, scrap or specific energy consumption. Captia Consulting ties each KPI to the decision it unlocks to avoid decorative dashboards.
Operational maturity: Level at which a plant’s processes, data and decisions are integrated, measured and continuously improved. Captia Consulting uses a maturity model to set the starting point and the next realistic step.
Operational SLA: Service-level agreement with realistic metrics, windows and penalties for industrial operation (response time, resolution, availability). Captia Service sizes them against historical data, not template values.
OT/IT: OT (Operational Technology) covers systems that control physical processes (PLC, SCADA, sensors). IT (Information Technology) covers corporate systems (ERP, CRM, data platforms). OT/IT convergence is the bridge on which Captia Connect builds ingestion and control.
Peak shaving: Technique that shaves contracted power peaks by shifting non-critical loads, using storage or activating self-consumption. Captia Energy coordinates it through the Connect system to avoid tariff penalties.
Physical AI: Physical AI is AI that perceives, reasons and acts on the real world. In a factory it starts by capturing good data from the equipment that already exists. It ranges from systems that read sensors and decide under supervision to robots, vision and autonomous agents. Each level rests on the previous one: without connected, comparable data with context there is no reliable perception to reason or act on.
Physical AI Ladder: The Physical AI Ladder is the Captia framework for placing a plant by what it can do with its data, not by the technology it buys. It has five levels: 0 Isolated, data lives in each machine; 1 Connected, equipment speaks a common language; 2 Aware, the plant knows what is happening in real time; 3 Acting, the system decides and executes under supervision; 4 Physical, autonomous perception and action. Each level requires the previous one.
PLC: Programmable Logic Controller, the industrial controller that runs control logic over sensors and actuators. Captia Connect treats it as untouchable: we integrate through standard protocols and never rewrite its safety logic.
Power Automate: Microsoft tool for automating workflows across apps (M365, Teams, SharePoint, third-party). Captia Service uses it as the automation fabric inside the customer Microsoft perimeter.
Predictive maintenance: Strategy that anticipates equipment failures using models over vibration, temperature or consumption data. Captia AI applies it over data ingested by Connect and turns alerts into executable work orders through Service.
PROFIBUS: PROFIBUS is the serial fieldbus that preceded PROFINET, maintained by PROFIBUS & PROFINET International and specified in IEC 61158 and IEC 61784. Its DP variant runs over shielded twisted-pair RS-485 at 9.6 kbit/s to 12 Mbit/s, with a master cyclically polling up to 126 addresses spread across segments that are terminated at both ends.
PROFINET: PROFINET is the industrial Ethernet standard maintained by PROFIBUS & PROFINET International (PI) and specified in IEC 61158 and IEC 61784. It carries cyclic process data between a controller and its field devices over standard Ethernet cabling, alongside conventional TCP/IP traffic. It defines a real-time class (RT) for millisecond cycles and an isochronous class (IRT) for synchronised motion.
Remaining useful life (RUL): Estimated time or cycles before an asset fails, produced by a model trained on vibration, temperature and consumption history. Captia AI turns it into an executable maintenance plan, not a mere number.
RPA: Robotic Process Automation, the automation of repetitive UI-based tasks across applications without decent APIs. Captia Service uses it as a tactical bridge, never as a replacement for real integration.
SCADA: Supervisory Control and Data Acquisition, a real-time supervision and data acquisition system for industrial processes. Captia Connect coexists with the existing SCADA instead of replacing it.
Scopes 1, 2 and 3: GHG Protocol categories to classify emissions: direct (1), indirect from electricity (2) and the rest of the value chain (3). Captia Energy calculates them from real operational data rather than generic estimates.
Sim-to-real gap: The sim-to-real gap is the difference in behaviour of a policy between the simulator it was trained in and the world it is deployed in. It is not only a matter of visual realism: it is caused mainly by poorly identified dynamic parameters, real sensor noise and latency, and contact. It is reduced by randomising the domain, identifying the system with measured data and fine-tuning with real episodes.
Smart factory: A smart factory is a plant whose equipment, systems and people share real-time data to continuously adjust production. It combines sensing, OT/IT connectivity and analytics on a common data architecture. Captia approaches it incrementally: connectivity and data context first, then concrete use cases built on that foundation.
Sparkplug B: Sparkplug B is the Eclipse Foundation specification that adds to MQTT what the base protocol leaves undefined: the shape of a message and its meaning. It fixes the topic hierarchy, types the payload with Protocol Buffers, and declares the state of every device through birth and death messages.
Specific energy consumption: Energy consumed per unit produced (kWh/t, kWh/m², kWh/unit). Captia Energy uses it as the primary metric because it normalises seasonality and volume and exposes real process inefficiency.
Transformation roadmap: Sequenced plan with waves, dependencies and success criteria that takes an industrial organisation from its current state to its target architecture. Captia Consulting always delivers it tied to capex/opex and measurable KPIs.
Unified Namespace (UNS): Industrial data architecture proposed by Walker Reynolds where a single pub/sub structure (typically MQTT) acts as the "single source of truth" for plant and business. Captia Connect uses it as the reference OT/IT pattern.
Vision-language-action policy: A vision-language-action policy takes visual observations and a natural language instruction and directly emits low-level actions such as pose increments or joint commands. It generalises to task variants described in words rather than trajectories programmed one by one, but its output is probabilistic: below it a deterministic safety envelope that does not depend on the model is still required.
WMS: Warehouse Management System, the software that governs receiving, slotting, picking and shipping. Captia Service implements it connected to industrial mobile terminals and the MES to close the plant-warehouse loop.
World model: A world model learns or encodes the dynamics of an environment: given a state and an action, what the next state is and what observation is expected, with its uncertainty. Unlike a predictive model, it allows actions that were never executed to be compared, planning inside the model and evaluating a policy without touching the plant. It requires a history of actions, not only of observations.
Zero Trust in OT: Security approach that removes implicit trust from flat plant networks: every access is authenticated, authorised and segmented. Captia Connect introduces it progressively over the existing industrial network.