Chapter 5: Actuators
The science exhibition was two days away, and the smart dustbin was not working.
Riya’s team had built it carefully. A sensor on the lid measured how far away a hand was. A small red light came on the moment a hand came close. The wiring was neat, the light blinked exactly on time, and the code did everything the team had planned.
And still the lid sat there, shut.
Their teacher looked at the bin for a while and asked one question. “The dustbin knows a hand is there. Now, what is going to lift the lid?”
Nobody had an answer, because nothing in the bin could push, pull or turn anything. Sensing is only half a machine. A dustbin that opens, a gate that swings, a lift that rises, a robot arm that picks up a cup — every one of them needs a part whose job is to produce movement.
Diagram
Fig. 5.0-1 – a dustbin with a closed lid; a hand held above it, a small sensor on the front of the lid with dashed lines going up to the hand, a lit red LED beside the sensor, and a large question mark drawn at the lid hinge where a moving part is missing
This chapter will answer each of these questions.
5.1 Why a Robot Needs Something That Moves
Think about what any machine does for you. You give it something, it gives you something back. What you give it is called the input, and what it gives back is called the output. A calculator takes numbers as input and gives numbers as output.
Now compare that with a dustbin whose lid should open by itself. Its input is a hand coming near. But its output cannot be a number on a screen. Its output has to be a lid that is actually up in the air.
That is the difference between a computer and a robot. A computer finishes its job when it produces the right information. A robot has not finished until something in the real world has physically changed position.
Sense, process, act
A robot does its work in three stages, and it needs a different kind of part for each one (see Fig. 5.1-1).
Diagram
Fig. 5.1-1 – three blocks joined left to right by arrows, labelled “Sense” (small sensor symbol), “Process” (small controller board symbol) and “Act” (small motor symbol); above the arrows, the words “information”, “information” and “movement” respectively
First the robot senses. A sensor measures something about the world — distance, light, temperature — and turns it into an electrical signal.
Then it processes. A controller reads that signal and works out, by following the rules stored in its program, what should happen next.
Finally it acts. Some part of the robot pushes, pulls, lifts or turns. Only in this last stage does the world outside the robot change at all.
Why a signal alone cannot lift a lid
It is easy to assume that once the controller has worked out the answer, the movement will somehow follow. It will not.
The signal coming out of a controller is a very small amount of electrical energy. It is enough to be measured and passed along a wire, but nowhere near enough to raise a lid or turn a wheel. Movement needs force, and force needs real energy — from a battery, a mains-powered supply, compressed air or pressurised oil.
So the third stage of a robot needs a part with a very particular job. It has to take a small command signal from the controller, draw energy from a supply, and use that energy to produce a push, a pull or a turn of exactly the size the controller asked for.
You already meet parts like this every day without thinking about them. The motor that spins a washing machine drum. The mechanism that slides a lift door open. The small rollers that pull a sheet of paper through a printer. The arm that raises the barrier at a parking gate.
Each of these takes an electrical command and gives back movement. That is a whole family of components with a name of its own, and the next section introduces it.
5.2 What an Actuator Is
The part that produces movement is called an actuator. Every lid that lifts, every wheel that turns and every robot arm that bends has one, and usually several.
The word comes from “act”. A sensor lets a machine sense; an actuator lets it act. Once you know the word, you start noticing them everywhere — in a printer, a lift, an automatic gate, a drone, even in the small vibrating unit inside a phone.
Two parts of that definition matter equally. The actuator must move something, and it must move it because a signal told it to. A spring that snaps back on its own is not an actuator, because nothing commanded it.
5.2.1 Signal In, Movement Out
Think of a tap in your school corridor. Your hand supplies only a small twist, yet a heavy stream of water comes out. Your hand does not supply the water. The water is already there, under pressure, and the small twist decides how much of it flows.
An actuator works in the same spirit. The signal from the controller carries almost no energy. The real energy comes from somewhere else — a battery, a mains supply, a tank of compressed air, a pump pushing oil. The signal only decides how much of that energy is let through, and in which direction.
So an actuator is really an energy converter. It takes energy in one form and gives it back as movement:
- electrical energy, in a motor, becomes a turning shaft
- compressed air, in an air cylinder, becomes a sliding rod
- pressurised oil, in an oil cylinder, becomes a strong straight push
This is the exact opposite of what a sensor does. A sensor takes something happening in the world and turns it into a small electrical signal. An actuator takes a small electrical signal and turns it into something happening in the world. Table 5.1 sets the two side by side.
5.2.2 Where the Actuator Sits in a Robot
An actuator never works alone. It sits at a fixed place in the chain of parts that makes a robot behave the way it does (see Fig. 5.2-1).
Diagram
Fig. 5.2-1 – a block diagram in a loop: “Input (task given to the robot)” arrow into “Controller”, arrow into “Actuator”, arrow into “Robot part that moves”, arrow out to “Output (task done)”; a return arrow from the moving part through a block labelled “Sensor” back to the Controller, this return arrow labelled “feedback”
Read the diagram from the left. A task comes in. The controller works out what movement that task needs. The actuator produces that movement. Some part of the robot — a wheel, a joint, a gripper — is carried along by it, and the job gets done.
Now look at the arrow coming back along the bottom. A sensor measures what actually happened and reports it to the controller. If the arm has not turned far enough, the controller keeps the actuator running a little longer. This returning information is called feedback, and you will study control loops that use it in detail in the next chapter.
One honest limitation is worth noting right at the start. An actuator has no idea what it is doing. It cannot tell whether it is lifting a light plastic lid or crushing a finger; it simply converts as much energy as it is told to. Everything sensible about a robot’s movement comes from the sensors and the controller around the actuator, never from the actuator itself.
5.3 Types of Actuators: Linear and Rotary
Once you start looking at actuators, you find hundreds of different ones. Some are the size of a grain of rice; some are as thick as a tree trunk. To make sense of them, engineers group them.
The simplest and most useful grouping asks one question: what shape of movement does it produce? Ask that, and every actuator falls into one of two families — linear actuators, which move in a straight line, and rotary actuators, which turn.
This grouping is useful because it matches the way you think about a job. A gate has to swing, so you need something that turns. A drawer has to slide out, so you need something that pushes in a straight line. Choose the family first, and the choice of exact component becomes much easier.
5.3.1 Linear Actuators
Picture the door of a lift sliding open, or a drawer coming out of a desk. The moving part travels along a straight path, out and back, and nothing goes round in a circle.
The distance between the two ends of that travel has a name. It is called the stroke. A linear actuator with a stroke of 10 cm can push its rod out by 10 cm and no further, however long you leave the power on. This is the honest limit of the whole family: the movement always stops at the end of the stroke.
Linear actuators are used wherever something must be lifted, pressed, clamped or slid (see Fig. 5.3-1).
Diagram
Fig. 5.3-1 – a linear actuator drawn as a cylinder with a rod coming out of one end; the rod shown in two positions, retracted and extended, with a double-headed arrow between them labelled “stroke”; the far end of the rod attached to a hinged flap it is pushing open
5.3.2 Rotary Actuators
Now picture a ceiling fan, or the wheels of a bus. Here the moving part goes round and round, and it can keep going for as long as power is supplied.
Most rotary actuators are electric motors, and you will meet the main kinds in section 5.5. What matters here is the quantity that describes their strength. A rotary actuator does not push; it twists. The turning effect it produces is called torque, and it depends both on the force applied and on how far from the axis that force acts.
\[\text{torque} = \text{force} \times \text{distance from the axis}\]
Here force is measured in newtons (N) and the distance in metres (m), so torque is measured in newton metres (N m). This is why a spanner with a long handle loosens a tight bolt more easily than a short one — the same force, applied further from the axis, gives more torque.
A rotary actuator has one clear advantage over a linear one: its movement has no end. A shaft can turn a hundred times or a hundred thousand times. That is exactly what a wheel needs, and it is why a robot car can keep travelling as long as its battery lasts (see Fig. 5.3-2).
Diagram
Fig. 5.3-2 – a small DC motor with its shaft pointing right, a wheel fixed to the shaft, a curved arrow around the shaft labelled “rotary motion” and a straight arrow along the ground labelled “the robot moves forward”
The turning of the wheel and the travel of the robot are linked by simple geometry. Every full turn of the wheel carries the robot forward by one circumference.
5.3.3 Linear or Rotary: Choosing for the Job
The two families are not rivals. Each suits a different shape of task, as Table 5.2 shows.
Turning one kind into the other
Sometimes the actuator you have is not the shape you need. That is not a dead end, because a simple mechanism can change one kind of motion into the other (see Fig. 5.3-3).
A screw and nut turn rotation into a straight push — this is exactly what happens inside the electric linear actuator described earlier. A rack and pinion does the same thing with a small gear running along a toothed bar. And a wheel does the reverse in effect: the shaft turns, and the whole robot moves in a straight line.
Diagram
Fig. 5.3-3 – two small sketches side by side; on the left, a threaded screw with a nut on it, a curved arrow on the screw and a straight arrow on the nut; on the right, a round toothed gear meshing with a straight toothed bar, a curved arrow on the gear and a straight arrow on the bar, labelled “pinion” and “rack”
So the real question when you design a machine is not “which actuator do I own?” but “what shape of movement does this job need, and what is the simplest way to produce it?”
5.4 Actuators at Joints and Wheels
Knowing the two families of actuator is one thing. Knowing where they are fitted on a real robot is another.
In almost every robot you will ever build or see, the actuators sit in one of two places: at a joint, so that one part of the robot can move against another, or at a wheel, so that the whole robot can travel. These two places are worth studying closely, because between them they cover most of what robots actually do.
5.4.1 The Actuator in a Robot Joint
Hold your arm straight out and bend it at the elbow. Your upper arm did not change shape, and neither did your forearm. Only the place where they meet allowed movement.
A robot arm is built the same way. The stiff parts that do not bend are called links, and the place where two links meet and can move relative to each other is called a joint. A joint by itself is only a hinge or a slide; it has no power of its own. Put an actuator at the joint and the arm comes alive (see Fig. 5.4-1).
Diagram
Fig. 5.4-1 – a simple two-link robot arm standing on a base; the straight sections labelled “link”, the two pivot points labelled “joint”, and a small motor drawn at each pivot with a curved arrow showing the direction it turns
Joints come in two shapes, and they match the two families of actuator exactly. A revolute joint turns, like your elbow, so it needs a rotary actuator. A prismatic joint slides in and out, like a drawer, so it needs a linear actuator.
A joint actuator is usually asked for something quite precise: not “keep turning” but “turn to this angle and stay there”. A motor on its own cannot do that, because switching it off does not tell you where it stopped. So a joint actuator is normally built as a package — a motor, a set of gears to increase its torque, and a sensor that reports the angle reached, all in one small body. The controller keeps the motor running until the reported angle matches the angle asked for. Section 5.5 looks at the motors inside these packages.
5.4.2 The Actuator in a Robot Wheel
A wheeled robot has a simpler problem to solve. It does not need to stop at an exact angle; it needs to keep rolling. That is the natural job of a rotary actuator, so a small geared motor is fixed directly to each driven wheel.
Most classroom robots use two driven wheels, one on each side, with a small free-rolling wheel at the front or back that carries weight without being driven (see Fig. 5.4-2). This third wheel has no actuator at all. It only keeps the robot from tipping over.
Diagram
Fig. 5.4-2 – top view of a rectangular robot chassis; a motor and wheel on the left edge and another on the right edge, each with a curved arrow, and a small free-rolling wheel at the front centre labelled “not driven”
With two independent wheel actuators, the robot can do everything it needs without any steering mechanism at all:
- both wheels turning forward at the same speed — the robot goes straight
- the left wheel slower than the right — the robot curves to the left
- one wheel stopped and the other turning — the robot swings around the stopped wheel
- the two wheels turning in opposite directions — the robot spins on the spot
Notice what is doing the thinking here. The wheels do not know any of this. The controller sends a different command to each actuator, and the difference between the two commands is what produces the turn.
5.5 Actuators Grouped by What Powers Them
Linear and rotary tells you the shape of the movement. It does not tell you where the energy for that movement comes from, and that turns out to matter just as much.
A robot arm in a school lab runs off a battery. A digging machine at a building site runs off oil pushed by a pump. A gripper in a bottling plant runs off compressed air. All three are actuators, and all three could be producing the very same straight-line push — but they are built completely differently, because they are fed completely differently.
So there is a second way of grouping actuators: by the kind of energy they convert. The five kinds you should know are electric, pneumatic, hydraulic, piezoelectric and shape memory alloy actuators.
5.5.1 Electric Actuators
An electric actuator converts electrical energy into movement. This is by far the most common kind, and almost every actuator you will use at school is one.
Nearly all electric actuators are built around a motor, and four kinds of motor cover most of robotics (see Fig. 5.5-1).
Diagram
Fig. 5.5-1 – photo of three small motors laid side by side on a table: a plastic-geared DC motor with a wheel, a hobby servo motor with a white plastic horn, and a stepper motor with a visible metal body and coloured wires
A DC motor runs on direct current, the steady current a battery supplies, and turns continuously for as long as it is powered. Raise the voltage and it turns faster; swap the two connections and it turns the other way. It is cheap, simple and ideal for wheels.
An AC motor runs on alternating current, the kind that comes from a mains socket. These are the motors in fans, pumps and factory conveyor belts, where a large machine has to run steadily for hours.
A servo motor is a complete package: a DC motor, a set of gears, an angle sensor and a small control circuit, all in one body. You do not tell it to turn; you tell it an angle, and its own circuit runs the motor until the sensor reports that the angle has been reached. This is what makes it the standard choice for a robot joint. A typical hobby servo of the kind used in school kits can be commanded anywhere between 0° and 180°.
A stepper motor turns in fixed, equal steps rather than smoothly. Each pulse sent to it advances the shaft by one step, so counting pulses tells you exactly how far it has turned, with no sensor needed at all.
The step size is fixed by how the motor is built, and it makes the arithmetic of movement very easy.
Electric actuators are precise, available in every size from a fingertip to a car engine, and easy to connect to an electronic controller. Their honest limit is force: for really heavy lifting, an electric actuator becomes large, hot and expensive compared with the alternatives you will meet next.
5.5.2 Pneumatic Actuators
Think of a bicycle pump with your thumb over the outlet. Press hard and the air inside squashes; let go and it springs back. Air can store energy by being squeezed, and that stored energy can be released to push something.
A pneumatic actuator uses compressed air to produce linear or rotational motion. The usual form is a cylinder with a piston inside it and a rod attached to the piston (see Fig. 5.5-2). Its working can be described in four steps.
Diagram
Fig. 5.5-2 – a cut-away cylinder with a piston in the middle and a rod coming out to the right; a control valve on the left with an air supply line entering it and an exhaust port beside it; arrows showing air entering behind the piston and the rod moving out
Compressed air supply: compressed air is supplied to the actuator through a control valve.
Air intake and piston movement: when the control valve is opened, compressed air flows into the cylinder on one side of the piston, and the pressure of that air forces the piston to move in a straight line.
Linear motion: as the piston moves, it carries the rod with it, and this straight-line motion does the useful work — lifting, pushing or pulling.
Exhaust and return stroke: when the piston reaches the end of its stroke, the control valve switches the air supply to the other side of the piston. The air on the first side escapes through the exhaust port, and the piston is pushed back to where it started.
Pneumatic actuators are quick, because air moves fast through a pipe. They are simple to build and easy to maintain. And because there is no electricity at the moving end, they can be used in places where a spark would be dangerous, such as a paint shop or a flour mill.
They are common in pick-and-place machines, sorting and packing lines and assembly lines, where the same short push is repeated thousands of times a day.
Their weakness comes from the same property that makes them useful. Air squashes. That means the rod does not stop exactly where you want it to, and it gives a little when the load pushes back. So pneumatic actuators are excellent at going all the way out and all the way back, and poor at stopping precisely in the middle.
5.5.3 Hydraulic Actuators
Now replace the air with oil. Oil hardly squashes at all, so pressure applied at one end is passed on almost completely to the piston at the other end. That single difference changes what the actuator can do.
A hydraulic actuator uses pressurised fluid to produce very high force and torque. A pump draws the fluid from a reservoir, raises its pressure and sends it through a control valve into the cylinder. The fluid presses on the piston, which slides out in a straight line. When the piston reaches the end of its stroke, the valve switches the flow to the other side and the fluid from the first side returns to the reservoir, pushing the piston back (see Fig. 5.5-3).
Diagram
Fig. 5.5-3 – a loop showing a reservoir tank, a pump drawing fluid from it, a pipe to a control valve, two pipes from the valve to the two ends of a cylinder with a piston and rod, and a return pipe from the valve back to the reservoir; arrows on the pipes showing the direction of flow
This is the family behind the machines that move real weight: excavators and other construction and mining equipment, heavy-duty industrial robots, aircraft landing gear, and the lift that raises a car in a service station. Their movement is smooth and controlled even under a very heavy load.
The price is complication. A hydraulic system needs a pump, a tank, valves and pipes, all of which add weight. Seals wear out, and a worn seal leaks oil, which is both a maintenance problem and a mess on the ground.
5.5.4 Piezoelectric Actuators
Some crystals do something surprising. Squeeze them and a small voltage appears across them. This is called the piezoelectric effect, and you can find it in a gas lighter, where pressing the button strikes a crystal and the voltage produced jumps across as a spark. Quartz and lead zirconate titanate are common piezoelectric materials.
The effect also works the other way round, and that is what an actuator uses. Apply a voltage across the crystal and it changes shape very slightly — it stretches or squeezes by a tiny amount. A piezoelectric actuator is built from this reversed effect: voltage in, a very small but very exact movement out (see Fig. 5.5-4).
Diagram
Fig. 5.5-4 – a small block of piezoelectric material between two metal plates connected to a voltage source; the block shown twice, once with the supply off and once with the supply on, the second slightly taller, with a magnified scale bar showing the tiny change
The movement is measured in fractions of a millimetre, which sounds useless until you meet a job that needs exactly that. Piezoelectric actuators position the sample under a powerful microscope, adjust optical instruments and precision machining tools, and push single drops of ink out of the nozzles of an inkjet printer. They are compact, light, and respond almost instantly.
Their limits are equally clear. The total movement is tiny, so they cannot lift or carry anything across a distance, and they usually need a high voltage to produce even that small displacement.
5.5.5 Shape Memory Alloy (SMA) Actuators
The last kind uses no motor, no pump and no moving joint at all. It uses a metal that remembers.
A shape memory alloy (SMA) is a metal that can be bent out of shape while cool, and returns to its original shape when it is heated. The best known is nitinol, an alloy of nickel and titanium. Heat a bent nitinol wire — usually just by passing a current through it, so that its own resistance warms it — and it pulls itself straight again. Let it cool and it can be stretched back ready for the next time (see Fig. 5.5-5).
Diagram
Fig. 5.5-5 – a thin nitinol wire shown in two states: on the left, cool and stretched long, holding a small lever down; on the right, with a current flowing through it, the wire shortened and the lever pulled up, with a small heat symbol beside it
An SMA actuator is startlingly strong for its weight, silent, and small enough to fit where nothing else would. That is why it appears in medical devices and in micro-robotics, where space is severely limited.
The catch is time and heat. Making the wire hot is quick, but waiting for it to cool again is not, so an SMA actuator cannot move rapidly back and forth. It also draws current the whole time it is holding its shortened shape, and the movement available is short.
5.5.6 The Five Kinds Compared
Each kind wins at something and loses at something else. Table 5.3 sets them side by side.
Read the last column carefully. It is the one that decides most real designs, because a component is usually rejected for what it cannot do rather than chosen for what it can.
5.6 The Role of Actuators in Robotics
So far the actuator has been described as the part that moves. That is true, but it undersells the job. Look at what a robot is actually asked to do in a workplace and you find the actuator carrying five separate responsibilities.
Motion generation
The first role is the obvious one. Actuators are the parts that make anything happen at all — everything else in the robot only handles information.
In a courier sorting centre, motors drive the long belts that carry parcels from one end of the building to the other. In a drone, motors spin the propellers that hold the whole machine in the air. Remove the actuators from either and you are left with a very well-informed pile of metal.
Precision control
The second role is about how the movement is produced, not whether it happens. A good actuator does not simply move; it moves by a controlled amount, at a controlled speed, and stops where it was told to.
A machine that places electronic components on a circuit board has to put each one down within a fraction of a millimetre of its correct position, several times a second. That accuracy is not a lucky accident. It comes from actuators that can be commanded to an exact position and can hold it.
Force application
The third role is strength. Many tasks are not about position at all but about pushing, pulling, gripping or pressing with enough force to get the job done.
A palletising robot in a grain warehouse lifts full sacks off a belt and stacks them on a pallet. A press in a small workshop forces a metal sheet into shape. Neither task needs delicate positioning; both need an actuator that will not give way under load.
Force cuts both ways, though. The same actuator that can lift a sack can also crush whatever is in its path, which is why heavy machines are fenced off, fitted with sensors, and stopped by an emergency switch that cuts the power to the actuators directly.
Adaptability and flexibility
The fourth role is variety. Because a robot can carry several actuators, each moving a different part in a different direction, one machine can be set up for many different tasks.
An arm with several joints can reach around an obstacle that would defeat a fixed machine. A soft gripper — one whose fingers are air-filled pads rather than rigid metal — closes around a mango or a tomato in a packing shed and takes its shape, holding it firmly without bruising it. Changing the task usually means changing the program and the tool, not rebuilding the robot.
Environmental interaction
The fifth role is what all the others add up to. Actuators are the only way a robot can change the world around it rather than merely observe it.
A floor-cleaning robot meets the leg of a table, reverses and sets off in a new direction. A driver-assisted car adjusts its steering, its acceleration and its braking in response to what its sensors report. In each case the sensors supply the information, the controller applies the rules in its program, and the actuators are what actually turn, brake or steer.
5.7 Choosing the Right Actuator for a Task
Suppose you are designing a machine and you have reached the moment of choosing its actuator. There is no single best one. There is only the one that fits this job, in this place, with the supply you have.
Beginners usually start from the wrong end — they pick the component first, because it is the one in the kit, and then try to make the task fit it. Experienced designers start from the task and let it narrow the choice down.
Five questions do most of the work.
Run the questions in order and the answer usually arrives before you reach the end (see Fig. 5.7-1).
Diagram
Fig. 5.7-1 – a flowchart starting from a box “What must move, and how?”; first branch “straight line” or “turning”; the straight-line branch leading to a question “very high force?” with “yes” going to “hydraulic” and “no” going to “electric or pneumatic”; the turning branch leading to a question “must it stop at an exact angle?” with “yes” going to “servo or stepper motor” and “no” going to “DC motor”; a small side box off the whole chart reading “very small, very fine movement? — piezoelectric or shape memory alloy”
Step 2 is the one students most often skip, and it is the one that decides whether the machine works at all. Torque was defined in section 5.3.2 as force multiplied by the distance from the axis, and that is all the arithmetic you need.
That figure is a floor, not a target. It is the torque needed merely to stop the arm sagging, with the link itself weighing nothing and nothing moving. A real arm has its own weight, and lifting takes more torque than holding. So an actuator is always chosen with a comfortable margin above the calculated value.
Table 5.4 shows the five questions applied to a few ordinary tasks.
5.8 Summary
- A robot works in three stages — sense, process, act. Sensors and the controller handle only information; nothing in the world changes until the acting stage.
- An actuator is a component that receives a control signal and uses energy from a supply to produce a physical movement, such as a push, a pull or a turn.
- The signal from the controller carries almost no energy. The actuator draws real energy from a battery, a mains supply, compressed air or pressurised fluid, and converts it into movement.
- A sensor works from the world into the robot; an actuator works from the robot out into the world.
- In a robot, the actuator sits between the controller and the part that moves, and a sensor usually reports the result back to the controller as feedback.
- By the shape of the movement they produce, actuators are of two kinds: linear actuators, which move in a straight line, and rotary actuators, which turn about an axis.
- The travel of a linear actuator is limited to its stroke. A rotary actuator can keep turning without limit.
- The turning effect of a rotary actuator is called torque, and it is equal to the force multiplied by the distance from the axis. It is measured in newton metres.
- Simple mechanisms such as a screw and nut, or a rack and pinion, change rotary motion into linear motion.
- The stiff parts of a robot arm are links, and the places where they meet and move are joints. A revolute joint turns and needs a rotary actuator; a prismatic joint slides and needs a linear actuator.
- A joint actuator is usually a package of motor, gears and an angle sensor, because a joint must stop at a set angle. A wheel actuator only has to keep turning.
- Two independently driven wheels can make a robot go straight, curve, swing round or spin on the spot, depending on the commands sent to each one.
- By the energy they convert, actuators are of five kinds: electric, pneumatic, hydraulic, piezoelectric and shape memory alloy actuators.
- Electric actuators use motors — DC motors for continuous turning, AC motors for mains-powered machines, servo motors for a commanded angle, and stepper motors for movement counted in fixed steps.
- Pneumatic actuators use compressed air. Their working has four steps: compressed air supply, air intake and piston movement, linear motion, and exhaust and return stroke. They are fast and simple, but air compresses, so they cannot stop precisely part-way.
- Hydraulic actuators use pressurised fluid, which hardly compresses, and so produce very high force. They need a pump, tank, valves and pipes, and their seals can leak.
- Piezoelectric actuators change shape slightly when a voltage is applied across the material. The movement is tiny but extremely precise and fast.
- Shape memory alloy actuators use a metal such as nitinol that returns to its original shape when heated. They are strong for their weight and very small, but slow to reset because the metal must cool.
- The roles of actuators in robotics are motion generation, precision control, force application, adaptability and flexibility, and environmental interaction.
- An actuator is chosen from the task: the shape of the movement, the force or torque needed, the speed, how exactly it must stop, the working conditions, and the supply available.
- Because an actuator acts on the real world, any machine that moves near people needs sensors that stop it and a way to cut power at once.
5.9 Key Terms
AC motor (alternating current motor) – A motor that runs on alternating current, the kind supplied by a mains socket, used in fans, pumps and factory conveyor belts.
Actuator – A component that receives a control signal and uses energy from a supply to produce a physical movement, such as a push, a pull or a turn.
DC motor (direct current motor) – A motor that runs on direct current, the steady current a battery supplies, and turns continuously while it is powered.
Electric actuator – An actuator that converts electrical energy into movement, almost always built around a motor.
Feedback – Information about what actually happened, measured by a sensor and sent back to the controller so that it can correct the movement.
Hydraulic actuator – An actuator that uses pressurised fluid to produce very high force and torque.
Input – What is given to a machine for it to work on.
Joint – The place where two links of a robot meet and can move relative to each other.
Linear actuator – An actuator that produces movement in a straight line, out and back along one direction.
Link – A stiff part of a robot arm that does not itself bend or change shape.
Output – What a machine gives back after it has worked on the input.
Piezoelectric actuator – An actuator that uses a material which changes shape slightly when a voltage is applied across it, giving a very small but very precise movement.
Piezoelectric effect – The property of certain materials, such as quartz, of producing a voltage when they are squeezed, and of changing shape slightly when a voltage is applied across them.
Pneumatic actuator – An actuator that uses compressed air to produce linear or rotational motion.
Prismatic joint – A joint whose two links slide in a straight line relative to each other, so it needs a linear actuator.
Revolute joint – A joint whose two links turn relative to each other, so it needs a rotary actuator.
Rotary actuator – An actuator that produces turning movement about an axis.
Servo motor – A package of motor, gears, angle sensor and control circuit that turns to a commanded angle and holds it.
Shape memory alloy (SMA) – A metal, such as nitinol, that can be bent out of shape while cool and returns to its original shape when it is heated.
Stepper motor – A rotary actuator whose shaft advances by a fixed angle each time it receives a pulse, so movement can be counted in steps.
Stroke – The distance between the two ends of the travel of a linear actuator.
Torque – The turning effect produced by a rotary actuator, equal to the force multiplied by the distance from the axis, measured in newton metres.
5.10 Objective Questions
The turning effect produced by a rotary actuator is called:
- stroke
- feedback
- torque
- pressure
The distance between the two ends of the travel of a linear actuator is called its ______.
State whether true or false: an actuator draws the energy for its movement from the control signal sent to it by the controller.
A joint whose two links slide in a straight line relative to each other is a:
- revolute joint
- prismatic joint
- rotary joint
- fixed joint
A machine has to work inside a flour mill, where an electrical spark would be dangerous. Which kind of actuator is best suited to it?
- electric
- pneumatic
- piezoelectric
- shape memory alloy
State whether true or false: a rotary actuator can go on turning for as long as it is supplied with energy.
Match each actuator in Column A with the energy it converts in Column B.
| Column A | Column B |
|---|---|
| a. Hydraulic actuator | i. Compressed air |
| b. Shape memory alloy actuator | ii. Electrical energy |
| c. Pneumatic actuator | iii. Heat |
| d. Electric actuator | iv. Pressurised fluid |
- A stepper motor has a step angle of 0.9°. How many steps does its shaft take to complete one full revolution?
5.11 Concept Questions
Define an actuator. State the two things it needs before it can produce any movement.
Give any three differences between a linear actuator and a rotary actuator.
A hydraulic actuator produces far more force than a pneumatic actuator of the same size. Give the reason.
A robot arm must stop at an exact angle and stay there. Explain why a servo motor is fitted at the joint instead of a plain DC motor.
Describe the working of a pneumatic actuator in four steps, naming each step.
A robot arm holds a 500 g object at the end of a link 20 cm long, with the joint at the other end of the link. Taking the acceleration due to gravity as 9.8 m/s² and ignoring the weight of the link, work out the torque the joint actuator must supply to hold the object steady. Show your working with units.
A stepper motor is used to move the head of a machine, and the controller counts the pulses it has sent to know where the head is. What would happen if the load became too heavy for the motor to complete one of its steps?
5.12 Case-Based Questions
- A school builds a trolley that carries books from the library to the classrooms. It runs on two wheels driven separately, one on each side, with a small free-rolling wheel at the front for balance. A flat lid on top has to lift to an angle of 90° so that books can be put in, and stay there until it is closed again.
- Name the kind of actuator you would fit to each driven wheel, and give one reason for your choice.
- Name the kind of actuator you would fit to the lid, and explain why the wheel actuator would not suit this job.
- When both wheels are commanded to turn at the same speed, the trolley drifts slowly to the left. Explain why this happens and what could be added to correct it.
- A fruit-packing shed is being fitted with a small automatic line. Three jobs on the line are listed in Table 5.5, along with what each one demands.
- For each of the three jobs, name a suitable kind of actuator and give the reason drawn from the table.
- The shed manager suggests using one hydraulic system for all three jobs, since it is the strongest. Give two reasons why this would be a poor choice.
- Workers stand beside the belt while the line runs. State two precautions the design must include because these machines move by themselves.