Craftsman Smart Factory flashcards
Craftsman Smart Factory · Build recall with a focused review session.
32 cards. Read the term and the definition here. The deck at the top of this page is optional.
- OEE formula
- Overall Equipment Effectiveness equals Availability times Performance times Quality. Availability is run time over planned production time; Performance compares actual output to ideal cycle time; Quality is good units over total units.
- Six Big Losses
- Breakdowns and setup/adjustment reduce Availability; minor stops and reduced speed reduce Performance; process defects and startup rejects reduce Quality. Mapping losses to these categories drives OEE improvement priorities.
- ISA-95 levels
- Five-level hierarchy: L0 process, L1 control, L2 supervision, L3 manufacturing operations (MES), L4 business planning (ERP). The L3-L4 boundary is the conventional IT/OT demarcation in smart factory architectures.
- Unified Namespace purpose
- A single ISA-95-aligned topic hierarchy where every asset publishes contextualized data over MQTT or OPC UA, replacing point-to-point integrations and giving all consumers one consistent, discoverable data model.
- Edge versus cloud rule
- Place processing at the edge when latency under roughly 100 ms, bandwidth cost, or data sovereignty dominate. Use cloud for long-horizon model training and cross-site correlation where seconds of latency are acceptable.
- OPC UA subscriptions
- Clients create monitored items with a sampling interval and a deadband; the server publishes notifications only when values change beyond the deadband, reducing traffic while preserving timestamps and quality flags.
- MQTT QoS levels
- QoS 0 delivers at most once, QoS 1 at least once with possible duplicates, QoS 2 exactly once via a four-way handshake. Higher QoS costs latency and broker state; choose per message criticality.
- Modbus register ranges
- Coils 0xxxx are read/write bits, discrete inputs 1xxxx read-only bits, input registers 3xxxx read-only 16-bit words, holding registers 4xxxx read/write 16-bit words. Values wider than 16 bits span registers.
- Sparkplug B payload
- An MQTT payload specification adding a topic namespace, birth and death certificates for session state, and stateful reporting where only changed metrics are transmitted, enabling interoperable SCADA integration.
- TSN time-aware shaping
- IEEE 802.1Qbv schedules transmission gates on a synchronized clock so time-critical frames get reserved windows on converged Ethernet. It requires 802.1AS time synchronization across all bridges.
- PLC scan cycle
- The repeating sequence of reading inputs, executing the control program, writing outputs, and housekeeping. Scan jitter, not just average scan time, determines whether fast control loops remain stable.
- PID output equation
- Controller output equals Kp times error plus Ki times the integral of error plus Kd times the derivative of error. Integral windup is handled by clamping or back-calculation; derivative acts on measurement to avoid setpoint kicks.
- Ziegler-Nichols tuning
- Raise proportional gain until sustained oscillation occurs, recording ultimate gain Ku and period Tu. Classic PID settings are Kp = 0.6 Ku, integral time Ti = Tu/2, derivative time Td = Tu/8.
- Safety performance level
- ISO 13849-1 assigns PL a through e based on risk graph inputs: severity of injury, frequency of exposure, and possibility of avoidance. The required PL determines architecture, diagnostics, and MTTFd targets.
- Safe torque off function
- STO removes drive power to the motor without contactor dropout, preventing torque production while the drive remains powered. It is a certified safety function independent of the standard control program.
- Robot singularity effect
- A configuration where two or more joint axes align, so the robot cannot achieve certain Cartesian motions. Near singularities joint velocities and torques grow unbounded, so paths must avoid them.
- Digital twin fidelity levels
- Descriptive twins model the as-is asset, predictive twins simulate future behavior, prescriptive twins close the loop with optimization. Each level requires an explicitly defined data synchronization direction and frequency.
- Digital thread definition
- The traceable lineage of data and decisions across an asset's lifecycle, from design through operation to service. It differs from a twin, which is a synchronized virtual instance of a specific asset.
- Asset Administration Shell
- The IEC 63278 standardized digital representation of an Industry 4.0 asset, composed of submodels such as identification, technical data, and documentation, exchanged as AASX packages or via APIs.
- Class imbalance metrics
- With rare defects, accuracy misleads. Use precision, recall, and their harmonic mean F1. Recall is critical when missing a defect is costly; precision matters when false alarms consume operator time.
- Bearing fault frequencies
- Vibration envelope analysis reveals characteristic frequencies: ball pass frequency outer race, inner race, and ball spin frequency, each modulated by shaft speed. Kurtosis and RMS track degradation severity.
- Remaining useful life validation
- RUL models must be validated on held-out assets, not random row splits, because degradation trajectories are correlated within an asset. Training requires run-to-failure histories or properly censored survival data.
- IEC 62443 zones and conduits
- Assets sharing security requirements are grouped into zones; all traffic between zones passes through controlled conduits. Each zone is assigned a target security level from SL1 to SL4.
- Defense in depth topology
- Traffic flows enterprise firewall to DMZ to cell firewall to controller. Historian mirrors and patch staging live in the DMZ; no direct enterprise-to-controller paths and no dual-homed engineering stations are permitted.
- Legacy protocol compensating controls
- Modbus and similar field protocols lack authentication and encryption. Compensate with network segmentation, allow-listed connections, one-way data diodes for outbound flows, and monitored conduits rather than direct exposure.
- OT incident response priority
- Protect the safe state of the physical process before rapid containment. Preserve forensic images of PLC logic and network captures before rebooting, since volatile controller state is lost on restart.
- Safety system independence
- A safety instrumented system must stay independent of the basic process control system and of cybersecurity remediation. Patching an SIS requires managed change control followed by revalidation of safety functions.
- Takt time formula
- Takt time equals available production time divided by customer demand. It sets the pace a line must meet; line balance efficiency compares the sum of task times against stations multiplied by cycle time.
- MTBF and MTTR
- Mean time between failures equals total operating time divided by number of failures. Mean time to repair equals total repair time divided by number of repairs. Both feed availability and maintenance strategy decisions.
- Event-driven integration threshold
- Point-to-point integration grows with the square of connected systems. When more than roughly five systems share data, move to a broker or publish/subscribe Unified Namespace to contain coupling and maintenance cost.
- Brownfield gateway mapping
- Legacy PLCs without native OPC UA connect through protocol gateways. Each tag must be mapped into the namespace hierarchy with a documented owner to prevent duplicate definitions and conflicting scaling.
- Smart factory project sequencing
- Asset inventory and network segmentation precede analytics deployment. Analytics built on unsynchronized or unreliable data produces results that cannot be reproduced or defended during acceptance testing.