Simplified teaching model
Structure + Control + Sensing + Actuation
Robotics Engineering Guide
Robots combine mechanical structures, motion systems, controllers, sensors, software, power systems and task-specific equipment into one interdependent architecture.
This guide explains the main components of a robot, how they work together, and which parameters and interfaces matter when selecting each subsystem.
Robot components are the mechanical, electrical, electronic and software subsystems that allow a robot to sense its state and environment, make control decisions and produce physical action.
Typical component groups include the mechanical structure, actuators, motors, transmissions, controllers, drives, sensors, machine vision, end effectors, power systems, communications and software.
The exact architecture depends on the robot’s function. An industrial arm, autonomous mobile robot, cobot, humanoid and inspection robot do not contain the same arrangement of components.
is the currently published international robotics-vocabulary standard, although ISO now identifies a replacement draft under development.
There is no universal number of robot components.
A basic functional model may describe structure, actuation, control and sensing. A detailed engineering model may separately identify motors, reducers, drives, controllers, sensors, vision, power, communication, software and end effectors.
The useful component taxonomy is the one that makes system functions, interfaces, design ownership and supply dependencies explicit.
Structure + Control + Sensing + Actuation
Motor + Drive + Reducer + Encoder + Brake + Joint
Mechanical + Motion + Control + Perception + Power + Software + Task equipment
An IFR robotics-programme summary has described reducers, robot motors, controllers, sensors and terminal or end-effector technologies as five core component areas. Use this only as one industry framework—not as a universal definition of every robot. IFR World Robotics R&D Programs summary.
Do not assume a fixed count. Every robot does not have exactly eight components. The number of identifiable subsystems depends on how the architecture is decomposed and which functions the robot must perform.
Modern robots are built from functional systems that span mechanics, motion, control, perception, power and software. The ten blocks below describe how engineering teams typically partition a robot architecture. Not every robot contains every element in equal depth.
An actuator converts electrical, hydraulic or pneumatic energy into controlled physical motion.
The motor produces torque and speed. It may operate directly or through a transmission and is normally controlled through a drive using position, velocity or torque feedback.
The transmission changes the relationship between motor speed and output torque and influences backlash, stiffness, accuracy, efficiency and joint size.
The control system coordinates motion, reads feedback, executes robot and application logic, manages communication and enforces defined safety functions.
Position, velocity, torque, current, temperature.
Distance, objects, people, contact, terrain and workpieces.
Machine vision converts optical information into measurements or decisions used for localisation, guidance, inspection, identification and interaction.
provides a standardized basis for objectively characterizing industrial cameras and image sensors.
The end effector is the interface between the robot and the task. Its mass, centre of gravity, inertia, utilities and hazards affect the complete robot system.
Hardware compatibility does not guarantee software or protocol compatibility.
These six component categories have dedicated engineering guides. Each guide covers parameters, interfaces and selection risks at component level. This hub explains how they fit into the wider robot architecture.
Function: Produce controlled rotary or linear movement.
Primary risk Nominal torque may be adequate while duty-cycle, stiffness or thermal performance is not.
Read the actuators guide GuideFunction: Provide compact high-ratio transmission for precision robot joints.
Primary risk Backlash alone does not describe transmission stiffness, accuracy or life.
Read the reducers guide GuideFunction: Generate controlled torque and speed under closed-loop control.
Primary risk A motor can meet peak torque while failing continuous thermal or inertia needs.
Read the servo motors guide GuideFunction: Coordinate robot motion, sensing, communication, safety and task logic.
Primary risk The hardware may be suitable while the software, protocol or licensing model is incompatible.
Read the controllers guide GuideFunction: Measure robot state, interaction and environment.
Primary risk A specification without test conditions may not predict real system performance.
Read the sensors guide GuideFunction: Detect, measure, localise and inspect through optical sensing.
Primary risk Selecting a camera before defining the scene, optical geometry and decision task.
Read the machine vision guide GuideEncoder / force / current feedback → Controller
A component should not be evaluated only through its individual datasheet.
Changing the reducer can affect motor inertia, controller tuning, joint stiffness, bearing loads and thermal behaviour. Increasing camera resolution can affect optics, lighting, processing load, bandwidth and latency.
Component selection is therefore a system-level engineering decision.
An articulated robot arm chains structural, motion, control and task subsystems from base to end effector. The sequence below is a simplified exploded view for engineering discussion—not a universal bill of materials.
| Arm subsystem | Primary function | Key sourcing consideration |
|---|---|---|
| Structure | Carry loads and maintain geometry | Stiffness, mass and fatigue |
| Motor | Generate torque and speed | Thermal limits and inertia |
| Reducer | Increase torque and reduce speed | Lost motion, stiffness and life |
| Bearings | Support radial, axial and moment loads | Load capacity and preload |
| Encoder | Measure position and motion | Resolution, accuracy and interface |
| Brake | Hold or stop defined axes | Holding torque and release behaviour |
| Drive | Control motor current and motion | Compatibility and tuning |
| Cabling | Carry power, signals and utilities | Flex life and routing |
| Controller | Coordinate axes and application | Cycle time, software and protocols |
| End effector | Perform the task | Payload, utilities and hazards |
No. Robots share functional needs such as motion, sensing, control and power, but implement them differently according to their task and environment.
| Robot type | Architecture emphasis |
|---|---|
| Industrial robot arm | Joint motors, precision reducers, encoders, controller and end effector |
| Collaborative robot | Integrated sensing, torque control, safety functions and accessible programming |
| AMR | Drive wheels, localisation sensors, batteries, navigation compute and fleet connectivity |
| Service robot | Mobility, perception, human interface, task module and remote support |
| Humanoid robot | High joint count, compact actuators, batteries, balance sensing and thermal control |
| Inspection robot | Special mobility, environmental protection, communication and sensor payload |
| Delta robot | Lightweight parallel structure, high-speed actuation and vision integration |
Component selection should begin with the robot’s function, loads, motion profile, operating environment and system interfaces—not with a component catalogue. The nine dimensions below frame a system-level review.
remains the current ISO performance-criteria and test-method standard for manipulating industrial robots after its confirmation in 2021. It applies at robot-system level and should not be treated as a substitute for component-specific test methods.
| Interface | Questions |
|---|---|
| Motor ↔ reducer | Shaft, flange, input speed, torque, inertia and lubrication |
| Reducer ↔ joint | Output support, moment load, alignment, stiffness and mounting |
| Motor ↔ drive | Voltage, current, feedback, commutation and tuning |
| Drive ↔ controller | Protocol, control cycle, synchronization and safety |
| Sensor ↔ controller | Signal, range, latency, bandwidth and calibration |
| Camera ↔ compute | Physical interface, driver, bandwidth and timing |
| Controller ↔ software | API, SDK, operating system, licensing and updates |
| End effector ↔ robot | Flange, payload, inertia, utilities and safety |
| Power ↔ system | Voltage, peak load, protection, grounding and EMC |
| Component ↔ environment | Temperature, ingress, contamination and vibration |
Two components with similar external dimensions are not necessarily interchangeable. Mechanical, electrical, control, software, thermal and lifecycle interfaces must all be considered.
| Dimension | Integrated module | Discrete architecture |
|---|---|---|
| Development speed | Usually faster | Usually slower |
| Internal integration | Supplier-controlled | Buyer-controlled |
| Customisation | Limited | Greater |
| Repair model | Module replacement | Component-level service possible |
| Architecture visibility | May be limited | Greater |
| Supplier lock-in | Generally higher | Potentially lower |
| Alternative sourcing | More difficult | Potentially easier |
| Engineering burden | Lower initially | Higher |
| Performance optimisation | Supplier-defined | Buyer-defined |
| Long-term cost | Depends on volume and dependency | Depends on engineering and scale |
Integrated modules are often suitable when development speed and known internal interfaces matter most. Discrete components are more appropriate when the buyer needs architecture control, performance optimisation, repairability or alternative sourcing.
| Mistake | Consequence |
|---|---|
| Selecting by nominal torque only | Thermal or duty-cycle failure |
| Comparing reducers only by backlash | Stiffness, life or lost-motion mismatch |
| Ignoring motor-load inertia | Poor control performance or instability |
| Selecting a camera before the optical task | Unusable images or excessive processing |
| Treating resolution as accuracy | Incorrect measurement assumptions |
| Ignoring software licences and SDK access | Integration or lifecycle lock-in |
| Assuming matching dimensions mean interchangeability | Electrical or control incompatibility |
| Ignoring cable and connector life | Intermittent field failures |
| Evaluating components separately | System-level performance failure |
| Selecting prototype parts without lifecycle review | Production and obsolescence risk |
| Accepting supplier substitutions without engineering review | Uncontrolled product change |
| Choosing lowest unit price | Higher integration, failure or redesign cost |
Supplier evaluation belongs primarily to commercial sourcing workflows. At guide level, engineers should verify six dimensions before treating a supplier as technically credible.
Who owns the design, firmware, tooling and critical subcomponents?
Can the supplier support interface definition, tuning and failure analysis?
Are production processes, end-of-line tests and reliability data documented?
Are measurement systems and calibration records maintained for critical parameters?
How are engineering changes, alternate parts and customer notification handled?
What spare-parts, firmware, repair and discontinuation commitments exist?
is scoped to ICT suppliers, but its supplier due-diligence model explicitly considers provenance, resilience, foundational cyber practices and supply-chain tiers. Those dimensions can inform reviews of connected controllers, sensors, cameras and other networked robotics components.
China offers broad robotics-component manufacturing capability, but supplier location and catalogue breadth do not establish component suitability. Buyers still need to verify system compatibility, product ownership, manufacturing processes, testing, critical sub-suppliers, software access, change control and lifecycle support.
China’s robotics-component ecosystem spans motion control and precision transmission, electronics and embedded-control manufacturing, sensors and machine-vision suppliers, and both standard and custom component production. A supplier may operate as an OEM, distributor or private-label brand, and critical subcomponents may be domestic or imported. Calibration, end-of-line testing, firmware ownership and export documentation vary widely between suppliers.
Supplier substitutions and undocumented engineering changes are a recurring risk in component programmes. Buyers should define acceptance tests, traceability requirements and notification rules before production release. Overseas support, repair access and software update rights should be confirmed for destination markets—not assumed from domestic sales capability.
Component sourcing in China is therefore an engineering verification problem as much as a supplier-search problem. Location alone does not reduce interface risk, software dependency or lifecycle exposure.
The main functional systems include mechanical structure, actuation, motors, transmissions, controllers, drives, sensors, machine vision, end effectors, power systems, communications and software. The exact arrangement depends on robot type and task. An industrial arm emphasises joint motors, reducers and a controller, while a mobile robot adds batteries, navigation sensors and fleet connectivity. See the functional systems section and the dedicated guides for actuators, servo motors and controllers.
There is no universal component count. A simplified model may describe structure, actuation, control and sensing, while a detailed engineering model separates motors, reducers, drives, controllers, sensors, vision, power, software and end effectors. The useful taxonomy is the one that makes interfaces and supply dependencies explicit. Do not treat “eight key components” lists as engineering definitions—they are teaching shortcuts, not specifications.
A robotic arm typically chains a base, structural links and joints, motors and drives, reducers, encoders, bearings, controller and cabling, a wrist and an end effector. Each joint repeats elements of the motion stack. The robotic arm components section lists subsystem functions and sourcing considerations. Detailed reducer and motor parameters belong in the harmonic reducers and servo motors guides.
A motor generates torque or rotational motion. An actuator is the broader motion-producing assembly and may include the motor, reducer, encoder, brake, bearings, housing and drive electronics. Selecting a motor without defining the transmission, feedback and thermal envelope does not define an actuator solution. Read the robot actuators guide for integrated-joint architecture and the servo motors guide for motor-level parameters.
Harmonic or strain-wave reducers provide high reduction ratios in a compact envelope with low backlash relative to many conventional gearboxes. That makes them common in precision robot joints where size, stiffness and accuracy matter. They are not the only transmission option—planetary, cycloidal and belt drives also appear in robotics. Compare lost motion, stiffness, efficiency and life in the harmonic reducers guide.
A robot controller coordinates motion, reads sensor feedback, executes application logic, manages communication and enforces safety functions. It may include or interface with servo drives, a real-time operating system and fieldbus connections. Hardware capability alone does not guarantee software, protocol or licensing compatibility. See the robot controllers guide for axis count, cycle time and integration parameters.
Robots use internal-state sensors such as encoders, current sensors and torque sensors, and external-environment sensors such as proximity, tactile, IMU and LiDAR devices. The required sensor set depends on control mode, safety architecture and perception tasks. Specifications without test conditions may not predict field performance. The robot sensors guide covers selection parameters and interface risks.
Machine vision converts optical information into measurements or decisions for localisation, guidance, inspection and interaction. It includes cameras, optics, lighting, processing hardware, software and calibration—not just the sensor module. Selecting a camera before defining scene geometry and the decision task is a common failure mode. Read the machine vision guide for resolution, latency and optical design considerations.
Robot subsystems communicate through electrical feedback interfaces, industrial Ethernet, EtherCAT, CAN, fieldbus links and software APIs. Power distribution, safety channels and real-time synchronization must be designed together. Hardware compatibility does not guarantee protocol or software compatibility. Controller and drive pairing should be verified before committing to a sensor or vision architecture.
Selection should start from robot function, loads, motion profile, environment and interfaces—not from catalogue browsing. Engineers evaluate mechanical, electrical, control, software, environmental and lifecycle requirements together. System-level performance cannot be inferred from isolated datasheets. Use the nine-dimension framework and the relevant child guide for each subsystem.
Sometimes, but interchangeability requires verified mechanical, electrical, control, software, thermal and lifecycle compatibility—not matching dimensions alone. Mixed-vendor architectures are common in custom robots but increase integration and validation burden. Interface matrices and sample testing reduce risk. Commercial sourcing support is described on Robotics Component Sourcing.
An integrated robot joint combines motor, reducer, encoder, brake, bearings and often drive electronics in one module. It can accelerate development when internal interfaces are proven, but may limit architecture visibility, repair options and alternative sourcing. Compare integrated modules against discrete architectures in the integrated versus discrete section and the actuators guide.
Start with a defined component requirement, interface specification and validation plan—not a broad supplier search. Verify product ownership, manufacturing and test capability, critical sub-suppliers, firmware access, change control and lifecycle support. China offers broad manufacturing depth, but location does not establish suitability. See the China sourcing section, supply-chain dependency guide and component sourcing service.
If you have defined the robot architecture and component requirements, Yana can help structure supplier research, interface comparison and validation planning for robotics components in China.