Process, Batch, and Discrete Automation

Introduction

Industrial automation applies sensing, computation, communication, and actuation to transform materials, energy, and information with limited direct human intervention. Mechatronics engineers encounter three major production patterns: continuous processes, batch processes, and discrete manufacturing. Many modern plants combine them in hybrid systems—for example, a food plant may continuously generate steam, execute recipes in batches, and package products as discrete units.

Correctly identifying the process type is a design decision, not merely a classification exercise. It influences instrumentation, controller selection, sequence design, alarm philosophy, data structures, safety analysis, production scheduling, and operator interfaces. This chapter develops practical methods for distinguishing these systems, writing control narratives, designing state-based sequences, applying ISA-88 concepts, and integrating machines with material-handling systems.

Learning Objectives

By the end of this chapter, the reader should be able to:

1. Distinguish continuous, batch, discrete, and hybrid processes.
2. Translate production requirements into a control narrative.
3. Design a state-based sequence with permissives, interlocks, transitions, and fault recovery.
4. Explain the ISA-88 batch-control model and recipe hierarchy.
5. Apply machine-control and material-handling principles to an industrial cell.
6. Evaluate safety, data, cybersecurity, and maintainability implications.

1. Classes of Industrial Processes

Continuous processes

A continuous process produces material whose flow is substantially uninterrupted over an operating period. Examples include oil refining, water treatment, cement production, and electricity generation. Typical controlled variables are temperature, pressure, flow, level, composition, and speed. Feedback control is central because disturbances occur while the process is operating.

A simple material-balance model for a tank is:

Here, V is liquid volume in m³, q_in and q_out are inlet and outlet volumetric flow rates in m³/s, and time is in seconds. The model assumes incompressible liquid and consistent flow units. If the flows are equal, volume is constant; if inlet flow exceeds outlet flow, level rises. A level controller may manipulate an outlet valve or pump speed.

Continuous systems are often represented by process-control diagrams and governed by PID loops, regulatory controls, advanced control, and safety instrumented functions. Stopping a process may be costly or hazardous, so start-up, shutdown, and abnormal-operation procedures require special attention.

Batch processes

A batch process makes a finite quantity of product according to a defined recipe. The product passes through identifiable stages such as charging, heating, mixing, reaction, holding, testing, and discharge. Pharmaceutical formulation, brewing, polymer production, and food processing are typical examples.

Unlike a continuous process, a batch has a beginning, an execution history, and an end. Its identity and genealogy matter: materials, equipment, operators, parameters, and deviations may need to be recorded. A batch can contain continuous control loops—for example, temperature control during a heating step—while its overall operation is sequential.

Discrete processes

A discrete process produces countable units. Examples include assembly, machining, filling individual containers, palletizing, and electronic-product manufacture. Sensors detect presence, position, orientation, and quality; actuators include cylinders, servomotors, grippers, conveyors, and robots.

Discrete automation is commonly organized around machine cycles. A cycle may include load, clamp, process, inspect, unload, and reset. Programmable logic controllers (PLCs), motion controllers, industrial robots, vision systems, and safety controllers are frequently integrated.

Hybrid processes

A hybrid system combines process types. A beverage facility may continuously control water treatment, execute syrup preparation in batches, and use discrete conveyors and robots for packaging. Hybrid architecture must preserve the requirements of each domain: regulatory control for continuous variables, recipe and genealogy management for batches, and deterministic sequencing for machines.

The distinction is operational rather than absolute. A batch reactor contains continuous variables, and a discrete packaging line may use continuous conveyor motion. Classification should therefore be based on the dominant production logic and information requirements.

2. From Production Requirement to Control Narrative

A control narrative is a structured description of what the system must do, under what conditions, and how it responds to abnormal situations. It forms a bridge between process engineering, electrical design, software development, commissioning, operations, and safety engineering.

A sound narrative identifies:

• purpose and operating modes;
• equipment and controlled variables;
• normal sequence of operation;
• permissives required before starting;
• interlocks that stop or inhibit actions;
• alarms and operator responses;
• failure behavior and recovery;
• manual, automatic, maintenance, and emergency modes;
• records and production-quality requirements.

A useful narrative distinguishes four related concepts. A permissive is a condition that must be true before an action is allowed. An interlock prevents or interrupts an action when an unsafe or damaging condition exists. An alarm informs an operator that attention is required but does not necessarily take automatic action. A trip commands a protective shutdown.

For example, a pump-start permissive may require sufficient tank level, an open discharge path, healthy motor protection, and no emergency-stop condition. A low-low tank level may be an interlock that stops the pump. A high motor temperature may generate an alarm before a protective trip occurs. The exact classification must be established by hazard and risk analysis rather than by convention.

Operating modes

A robust design defines mode authority explicitly. Common modes are:

• Off: outputs are placed in a safe state.
• Manual: an authorized operator commands individual devices, subject to safety interlocks.
• Automatic: the controller executes the approved sequence.
• Maintenance or setup: limited movement is permitted under controlled conditions.
• Emergency stop: hazardous motion or energy is removed according to the safety design.

Mode changes should not create unexpected motion. A transition from manual to automatic normally requires a state-consistency check, confirmation of permissives, and a defined takeover procedure.

3. State-Based Sequence Design

A state model represents the equipment or process as a finite set of operational states, events, and transitions. It is clearer and more testable than a long, unstructured list of Boolean conditions.

Typical states include:

1. Idle: ready but not operating.
2. Starting: outputs are being initialized and permissives verified.
3. Running: normal production actions are active.
4. Holding or Paused: operation is temporarily suspended without losing context.
5. Completing: the current unit or batch is being finalized.
6. Faulted: a fault has occurred and automatic progression is inhibited.
7. Aborting: the system is moving to a predefined safe condition.
8. Complete: production has ended and results can be reported.

A transition should have a clearly defined trigger and guard condition. For example:

Idle → Starting when Start is requested and all start permissives are true.

Starting → Running when initialization is complete and feedback confirms the required actuator states.

Running → Complete when the quantity target is reached and final inspection passes.

Running → Faulted when a monitored fault occurs.

The arrow notation describes logical movement, not a mathematical differential equation. A state should specify entry actions, continuous actions, exit actions, timeout behavior, and recovery behavior. A timer must be associated with a defined clock and unit; for instance, a heating phase may require temperature to remain above its target for 120 s, rather than merely reaching the target momentarily.

Sequence design principles

Single responsibility: each state performs a coherent operation. Avoid states that simultaneously perform unrelated tasks.

Explicit completion: every automatic step needs a completion condition, timeout, or intentional indefinite-hold designation.

Safe failure: sensor disagreement, communication loss, actuator failure, and loss of utility should produce a documented response.

Idempotence where possible: repeating a command should not create unintended consequences. For example, a valve-open command should remain safe if issued again.

Traceability: every transition and output should be linked to a requirement or approved operating procedure.

Example: indexing-and-drilling cell

An illustrative discrete cell receives a part on a conveyor, clamps it, drills a hole, checks the result, and releases the part. A simplified sequence is:

The controller should not infer that a part is clamped merely because a clamp output was energized. Feedback, such as a pressure switch or position sensor, is required where the risk and process quality justify it.

4. ISA-88 and Batch Control

ISA-88, commonly associated with batch-control models, separates the physical equipment from the procedural logic. This separation improves recipe portability, reuse, testing, and change control.

Physical model

The physical hierarchy is commonly described as:

• Enterprise: the business organization;
• Site: a plant location;
• Area: a major production grouping;
• Process cell: equipment that performs a batch process;
• Unit: an independent processing entity, such as a reactor;
• Equipment module: a functional grouping, such as a heating system;
• Control module: the smallest control entity, such as a valve, motor, or sensor loop.

Not every implementation uses every level explicitly, but the hierarchy provides a common vocabulary.

Procedural model

The procedural hierarchy is:

• Procedure: complete strategy for making a product;
• Unit procedure: activity performed by one unit;
• Operation: major processing activity, such as charging or heating;
• Phase: the smallest procedural element that commands equipment, such as open valve, dose material, or hold temperature.

A phase should expose parameters and status rather than embedding product-specific values in PLC logic. For example, a generic “Heat” phase may receive a target temperature, ramp limit, hold time, and maximum timeout from a recipe.

Recipes

A recipe defines how a product is made. A practical recipe may include material identities, quantities, equipment requirements, setpoints, order of phases, tolerances, sampling instructions, and reporting fields. Product-specific information should be separated from reusable equipment procedures.

A useful distinction is:

• Master recipe: approved generic definition of manufacture.
• Control recipe: a batch-specific execution instance derived from the master recipe.
• Equipment or site recipe: adaptation to available equipment and local constraints.

Recipe governance requires version control, authorization, electronic records where applicable, audit trails, and validation of changes. A setpoint is not automatically safe merely because it appears in a recipe; limits and interlocks must remain enforced by the control and safety systems.

5. Material Handling and Machine Integration

Material handling connects process equipment, machines, storage, inspection, and shipping. Typical devices include conveyors, elevators, automated storage and retrieval systems, palletizers, autonomous mobile robots, transfer carts, and robotic arms.

A material-handling design must address:

• flow capacity and buffering;
• routing and identification;
• accumulation and starvation/blocking;
• collision avoidance;
• safe zones and access control;
• part tracking and genealogy;
• recovery after jams or power loss;
• interfaces between line controllers and supervisory systems.

A conveyor line is not simply a series of motors. If the downstream machine stops, upstream equipment may need to stop, accumulate product, or divert it. The choice depends on product sensitivity, buffer capacity, cycle time, and safety constraints.

For a steady line, throughput is often limited by the bottleneck. If station cycle times are T₁, T₂, …, T_n seconds per unit and stations cannot be parallelized, the ideal line rate is bounded by:

where R_max is units per second and T_i is the effective cycle time of station i. This is an ideal capacity relationship: it assumes adequate material, negligible failures and changeovers, and no additional blocking or starvation. Real throughput is lower because of downtime, quality losses, and variability.

Overall equipment effectiveness is commonly expressed as:

Availability, performance, and quality are dimensionless fractions, normally between 0 and 1. Availability may be calculated as operating time divided by planned production time; performance compares actual rate with ideal rate; quality compares good units with total units. Definitions must be agreed because different organizations include changeovers, minor stops, or planned maintenance differently.

6. Applied Case: Illustrative Ready-to-Drink Beverage Line

Background

Consider a fictional beverage plant that prepares flavored syrup in batches, continuously supplies treated water, and fills bottles on a discrete packaging line. The case is illustrative rather than a report of a particular company.

Problem

The plant experiences inconsistent flavor concentration, frequent conveyor stoppages, and incomplete production records. Operators manually adjust recipes and restart machines after faults without a consistent recovery method.

Analysis

The system is hybrid. Syrup preparation requires batch phases and recipe control. Water treatment requires continuous flow, level, and quality control. Filling, labeling, case packing, and palletizing are discrete operations. The main risks are recipe errors, incompatible equipment states, loss of product identity, and unsafe restart after a jam.

Approach and solution

The engineering team defines an ISA-88 model for syrup production. A master recipe specifies ingredient identities, target masses, mixing speed, temperature limits, and hold time. Each batch receives a control recipe with a unique identifier. Charging, mixing, heating, sampling, and transfer are implemented as reusable phases.

The packaging line is redesigned as a state model: Ready, Load, Fill, Cap, Inspect, Reject, Pack, and Faulted. Each state has explicit sensor confirmation and timeout logic. A jam causes controlled stopping, displays the affected zone, records the event, and requires an authorized reset after the guard and safety conditions are restored. Restart resumes only from a validated state, not from an assumed sequence position.

A supervisory system records batch identity, tank identity, recipe version, operator actions, deviations, and packaging counts. Interfaces use defined data contracts: equipment state, command, permissive, fault, material identity, and timestamp. Safety functions remain independent of ordinary production logic where required by the risk assessment.

Outcome

The redesigned architecture improves repeatability and traceability, reduces ambiguous restarts, and makes recipe changes subject to approval. It does not eliminate all failures: sensors still require calibration, mechanical jams still occur, and production data must be reviewed for accuracy.

Learning points

1. A single factory can contain all process classes.
2. Batch recipes should not be confused with hard-coded sequence logic.
3. State models make abnormal behavior and recovery explicit.
4. Material tracking is a control requirement as well as a business requirement.
5. Safety, quality, and productivity must be engineered together.

7. Safety, Governance, and Contemporary Practice

Automation design must comply with applicable machinery, electrical, process-safety, and functional-safety requirements. Standards and regulations vary by jurisdiction and industry; engineers must consult current, authoritative editions rather than relying on informal summaries. Risk assessment should identify hazards from motion, pressure, temperature, chemicals, stored energy, unexpected restart, and human-machine interaction.

Safety functions should be specified with a required risk-reduction performance and validated independently. An ordinary PLC permissive may improve operational protection but may not satisfy a required safety function. Guard interlocks, emergency stops, safe torque off, light curtains, and safety instrumented functions must be selected according to the hazard and validated architecture.

Cybersecurity is part of availability and safety. Network segmentation, least privilege, authenticated access, controlled remote support, patch management, backups, and incident response reduce the chance that a cyber event becomes a production or safety event. Artificial intelligence can assist visual inspection, predictive maintenance, scheduling, and anomaly detection, but it should not silently replace validated protective logic. AI outputs require performance monitoring, representative data, change control, and a defined human response to uncertainty.

Good engineering practice includes simulation where feasible, version-controlled code, formal testing, alarm rationalization, calibrated instruments, clear naming conventions, operator training, and documented factory and site acceptance tests. A sequence is not complete when it works once; it is complete when normal, abnormal, maintenance, restart, and recovery behaviors have been tested.

Chapter Summary

Continuous processes produce flowing material and rely heavily on regulatory control and safety functions. Batch processes produce finite lots whose recipes, phases, history, and genealogy must be managed. Discrete processes produce countable units through machine cycles, motion, sensing, and inspection. Hybrid plants combine these characteristics and require coordinated architecture.

Control narratives translate production requirements into modes, actions, permissives, interlocks, alarms, trips, and recovery rules. State-based sequences provide a disciplined method for expressing normal and abnormal behavior. ISA-88 separates physical equipment from procedural control and supports reusable phases, controlled recipes, and traceable batch execution. Material-handling systems must address bottlenecks, buffering, routing, tracking, safety, and recovery. Effective automation integrates productivity with safety, quality, cybersecurity, maintainability, and responsible data governance.

Key Terms

Process automation: Use of measurement, control, and actuation to operate an industrial process.

Continuous process: Process producing flowing material with substantially uninterrupted operation.

Batch process: Process producing a finite quantity according to a recipe.

Discrete process: Process producing countable units through identifiable machine operations.

Hybrid process: System combining continuous, batch, and/or discrete characteristics.

Control narrative: Written description of operating logic, conditions, actions, and responses.

Permissive: Condition that must be true before an action is allowed.

Interlock: Logic that inhibits or interrupts operation under a specified condition.

Alarm: Notification requiring attention or response.

Trip: Protective command that places equipment or a process in a safer state.

State: Defined operating condition with associated actions and transitions.

Transition: Condition or event that causes movement from one state to another.

Phase: Smallest procedural batch element that commands equipment.

Recipe: Approved definition of materials, parameters, and procedures for making a product.

Genealogy: Record linking materials, equipment, processes, and products.

PLC: Programmable logic controller used for industrial monitoring and control.

Bottleneck: Resource that limits the capacity of a production system.

OEE: Product of availability, performance, and quality measures.

Safety function: Designed function that reduces risk to an acceptable level.

Review Questions

Short Answer Questions

1. What is the principal difference between a batch process and a continuous process?
2. Define a permissive and distinguish it from an interlock.
3. Name the main levels of the ISA-88 procedural model.
4. Why should a state have a timeout or explicit indefinite-hold designation?
5. What three factors are multiplied in the common OEE expression?

Long Answer Questions

1. Compare continuous, batch, discrete, and hybrid automation, including suitable examples and control challenges.
2. Explain how a control narrative is developed from a production requirement.
3. Describe the physical and procedural models of ISA-88 and explain their value in recipe-based production.
4. Discuss how a state-based sequence should handle faults, operator modes, and restart.
5. Explain the safety, cybersecurity, and governance issues that arise when connecting machines to supervisory systems.

Analytical / Application Questions

1. Design a state sequence for a robotic palletizer. Identify states, permissives, transitions, fault conditions, and recovery actions.
2. A line has effective station cycle times of 8 s, 10 s, 7 s, and 10 s per unit. Determine the ideal bottleneck-limited rate in units per minute and explain why actual production will usually be lower.
3. For a hybrid dairy plant, classify the water-treatment, pasteurization, filling, and palletizing operations. Propose an architecture that combines regulatory loops, batch recipes, discrete sequences, and material tracking.

Multiple Choice Questions with Answers

Question 1

Which characteristic most directly identifies a batch process?

A. It produces a finite lot according to a recipe.

B. It never uses feedback control.

C. It produces only countable mechanical parts.

D. It must operate continuously without interruption.

Correct Answer: A. It produces a finite lot according to a recipe.

Question 2

What is the primary purpose of a permissive?

A. To record historical production data

B. To allow an action only when required conditions are true

C. To replace every safety function

D. To increase motor speed automatically

Correct Answer: B. To allow an action only when required conditions are true

Question 3

Which item belongs to the ISA-88 procedural model?

A. Control module

B. Equipment module

C. Phase

D. Site

Correct Answer: C. Phase

Question 4

In a state-based sequence, what should normally happen when a critical fault occurs?

A. The sequence silently advances

B. The controller ignores sensor feedback

C. The system enters a defined fault or abort response

D. All alarms are disabled

Correct Answer: C. The system enters a defined fault or abort response

Question 5

A line has station cycle times of 8 s, 10 s, 7 s, and 10 s. What is its ideal bottleneck-limited rate?

A. 4 units/min

B. 6 units/min

C. 10 units/min

D. 12 units/min

Correct Answer: B. 6 units/min

Question 6

Which statement best describes a hybrid process?

A. It uses only manual control.

B. It combines different production-control characteristics in one system.

C. It cannot contain batch operations.

D. It produces no physical product.

Correct Answer: B. It combines different production-control characteristics in one system.

Question 7

Which item is most associated with batch genealogy?

A. Linking materials, equipment, parameters, and product history

B. Increasing conveyor acceleration

C. Removing all operator records

D. Replacing recipes with random setpoints

Correct Answer: A. Linking materials, equipment, parameters, and product history

Question 8

In the OEE expression, which factors are multiplied?

A. Pressure, temperature, and flow

B. Availability, performance, and quality

C. Speed, torque, and power

D. Cost, revenue, and profit

Correct Answer: B. Availability, performance, and quality

Question 9

Why is feedback confirmation important after energizing an actuator?

A. An output command alone does not prove that the physical action occurred.

B. It guarantees that no maintenance is required.

C. It removes the need for an emergency stop.

D. It makes every process continuous.

Correct Answer: A. An output command alone does not prove that the physical action occurred.

Question 10

Which practice best supports safe and maintainable automation?

A. Embedding all product recipes permanently in unversioned code

B. Testing only the normal production path

C. Using version control, documented tests, and defined recovery behavior

D. Allowing unrestricted remote access

Correct Answer: C. Using version control, documented tests, and defined recovery behavior