IELTS Academic Reading Practice Test 161

By the middle of the twentieth century, cars were capable of speeds that their brakes, tyres and bodies could not manage, and the number of people killed on the roads was rising every year. Engineers understood the problem but had no reliable way to study it. A collision lasts a fraction of a second, and the forces involved are far beyond anything a person can safely experience in a laboratory. Testing a new seat belt or dashboard on living people was out of the question; testing it on nothing at all, as the industry had largely done, meant guessing.

Those who first volunteered for controlled crashes were not car engineers but military researchers. John Paul Stapp, a physician in the American air force, was investigating what happens to pilots in crashes. At Holloman Air Force Base in New Mexico he built a rocket sled: a seat on rails, driven by rockets and stopped by a series of water brakes. In December 1954 Stapp himself rode it to 632 miles an hour and was brought to a halt in about a second and a half, enduring more than forty times the force of gravity. He survived, and his rides persuaded the air force to change the design of harnesses and seats.

Doctors had also begun to use a more disturbing source of data. At Wayne State University in Detroit, the biomechanist Lawrence Patrick and his colleagues ran impact experiments on cadavers, mostly unclaimed bodies from hospitals. The results gave engineers the first detailed measurements of how human bones, ribs and skulls respond to impact, and much of what is known about the limits of the body still rests on that work.

To repeat such tests on a large scale, engineers needed a substitute that was tough, standardised and could carry as many sensors as they wished. Among those who responded was Samuel Alderson, an American inventor whose company already made artificial limbs and medical instruments. His firm produced some of the first anthropomorphic test devices, as dummies are formally known: frameworks of steel and aluminium, covered with vinyl skin and joined by rubber parts that bend as human joints do. Inside the head, chest, pelvis and knees sit accelerometers and load cells, which record how violently each part of the body is thrown about and how much force passes through it.

The design that settled the matter came not from a university but from the research laboratory of General Motors. Its engineers built a dummy called Hybrid I in 1971, and a refined version, Hybrid II, was adopted for official tests in the United States in the early 1970s. Hybrid III, which appeared in 1976, is still the standard in many countries. It represents a man of about 1.75 metres and 78 kilograms, the fiftieth percentile of the population in the language of the industry, and versions scaled down to represent a small woman, a child and an infant followed.

Each dummy is a precision instrument, and each must be calibrated before it can be used. The head is dropped onto a rigid plate, the knee is struck by a swinging pendulum, and the results are compared with values established decades ago; a dummy that responds outside a narrow tolerance is not certified. In a crash test the dummy is seated, belted and wired to a recorder, and either the car is driven into a barrier or a sled carrying it is stopped within a fraction of a second. Several dummies may be needed for the driver and the passengers, and each must be examined again after the impact.

The weakness of the system is that dummies were shaped by the data available when they were designed, and those data came largely from the bodies of young servicemen and middle-aged men who had died in hospitals. The standard dummy is therefore male, and for decades the only smaller version in wide use was simply a scaled-down man. Women are not small men: their proportions, the distribution of their body mass and the strength of their necks differ. Analyses of real crashes have repeatedly found that women drivers are more likely to be seriously injured than men in comparable collisions.

Change is arriving slowly. In Sweden, the engineer Astrid Linder and her colleagues at the national road and transport research institute built a dummy modelled on the average woman, known as SET 50F, and presented it in 2022; it is intended for testing seats in rear impacts, where women appear to be at particular risk. Volvo, meanwhile, has drawn on its own archive of crashes since the 1970s, which includes both sexes and a wide range of body sizes, and has published much of the data for other manufacturers to use.

The dummy is now being joined, rather than replaced, by a different kind of model. Human body models are detailed computer simulations of bones, organs and tissues, built from medical scans and cadaver tests, and they can be varied in size, age and shape far more cheaply than any physical device. Some new cars are already designed partly against such simulations. But a simulation is only as good as the data used to check it, and some questions — how a muscle tenses in the instant before a crash, for example — can still be answered only by a physical test. The uncomplaining passenger is likely to be travelling for some time yet.

A maglev is a train supported, guided and driven by magnetic force, with nothing touching the track. Because nothing rubs, there is no rolling resistance and no mechanical wear, and the ride is quiet apart from the noise of the air itself. The lift can be produced in two quite different ways: magnets on the vehicle can pull it upwards towards a steel rail above them, or they can push it away from currents induced in coils in the guideway. The idea is old. The British engineer Eric Laithwaite demonstrated working models in the 1960s and worked on a hovertrain project that the government cancelled in 1973. Only Germany and Japan have since built full-scale high-speed systems.

The German system, marketed as Transrapid, uses attraction. The guideway is a T-shaped steel rail, and the train wraps around it, so that electromagnets on the vehicle sit beneath the edges of the rail. Current in the magnets lifts the train about ten millimetres clear and holds it there. The arrangement keeps most of the magnetic field inside the iron, but it is inherently unstable: if the gap widens, the pull weakens, and if it narrows, the pull strengthens, so that an uncontrolled train would be dragged into the rail. Sensors measure the gap several thousand times a second, and a control system adjusts the current in each magnet to keep it constant.

The Japanese approach uses repulsion. Superconducting coils on the train, cooled by liquid helium to within a few degrees of absolute zero, carry currents so large that they generate very strong magnetic fields. As the train moves, these fields sweep past coils set in the walls of the guideway, induce currents in them, and the magnetism that results pushes the train upwards. Because the effect depends on movement, the vehicle runs on retractable wheels until it is travelling at about 150 kilometres an hour. Once it lifts, it floats some ten centimetres above the track, a gap ten times wider than the German one and far more forgiving of small errors in construction.

Both systems are driven by a linear motor, which is an ordinary electric motor unrolled. In a conventional motor a rotor turns inside a ring of coils carrying alternating current; in a maglev the guideway carries the coils and the vehicle carries the magnets, so that a travelling magnetic wave drags the train along. Speed is set simply by the frequency of the current. With no gears and no engine, acceleration is smooth and strong, and the train can climb gradients of up to ten per cent, about twice as steep as a conventional railway. For braking, the motor can be reversed to work as a generator.

The speed record belongs to Japan. In April 2015 a seven-car train of the L0 series, on a test track in Yamanashi prefecture, reached 603 kilometres an hour. In commercial service the fastest is the Shanghai maglev, which links Pudong airport with the edge of the city: its 30-kilometre journey takes about seven minutes, at a top speed of 431 kilometres an hour. Slower urban lines also exist, among them Linimo, opened near Nagoya for the 2005 world exhibition, which runs at about 100 kilometres an hour. Japan's most ambitious line, the Chuo Shinkansen between Tokyo and Nagoya, began construction in 2014 and will run almost entirely in tunnel.

Supporters often claim that the absence of friction makes maglev exceptionally efficient, and at low speed that is true. At 500 kilometres an hour, however, almost all the resistance comes from the air, and the drag rises with the square of speed. The power required rises with the cube: doubling the speed takes roughly eight times as much power, and about four times as much energy for each kilometre travelled. Steel wheels could not run at such speeds anyway, but the advantage over high-speed rail narrows as speed rises. Emissions depend, of course, on how the electricity is generated.

Cost is the obvious obstacle. A maglev line cannot use existing track, so every kilometre must be new, and in populated country that means viaducts or tunnels; the Chuo Shinkansen is expected to cost tens of billions of dollars. The Japanese project also ran into a dispute with the governor of Shizuoka prefecture, who withheld approval for construction on the grounds that tunnelling would reduce the flow of water in a local river. Magnetic fields are a further concern, although the strength of a field falls rapidly with distance and operators measure it inside the cabin. Noise is another limitation: above about 300 kilometres an hour the dominant sound is generated by air passing over the body rather than by the magnets.

Research now runs in several directions. One aim is to replace liquid helium with cheaper cooling, or to use superconductors that work at higher temperatures, which would reduce the weight and cost of the equipment needed to keep the coils cold. Another is to remove the air altogether: the idea of running a levitated train through a tube pumped to a partial vacuum is more than a century old, and several companies are now trying to build one. For cities, the more promising market may be the low-speed maglev, which is quieter than a metro and can climb gradients that would defeat a tram.

On a wet evening, a box on six wheels hums along the pavement at walking pace, stops at a kerb and crosses. It carries no driver, only an insulated compartment holding someone's dinner. Such machines are now a familiar sight in a few places, and they raise an awkward question: should small self-driving delivery robots be allowed to use space designed for people on foot?

The technology is less mysterious than it looks. A typical robot runs on a battery and electric motors. Cameras facing in several directions, radar and ultrasonic sensors build a picture of its surroundings, while satellite positioning and a map tell it where it is. Software recognises people, bicycles and kerbs, and chooses a path at about six kilometres an hour. When the software is uncertain, a remote supervisor can watch over a number of machines at once and take control; the operator is not in the robot, but the robot is rarely alone.

Trials have moved from laboratories to ordinary streets. In Milton Keynes, a town in southern England with a network of redways, paths separated from most traffic, robots have delivered groceries and takeaway meals from local shops. University campuses have proved even more welcoming, because they are compact, mapped in detail and full of people who order food at short notice. Students at several universities in the United States and Britain have summoned a robot by phone and met it outside a hall of residence. The machine waits, opens a lid and requires a code before the compartment unlocks.

The case for allowing them rests on the last kilometre, the short journey between a shop or depot and a customer's door. A van that carries one hot meal across town burns fuel, occupies road space and needs a driver; a robot carrying the same meal from a local hub uses a fraction of the energy and can travel where a van cannot park. If deliveries are grouped at a hub, the number of van movements falls, and with them congestion and exhaust emissions. For retailers, the machines can extend opening hours without paying a night shift.

Against this, pavements are not empty corridors. They are the only safe place many people have to move about, and often barely wide enough for two people. A wheelchair user, a parent with a double pushchair or someone using a white cane needs a clear line and predictable behaviour from others. A robot that stops without warning, or waits in a row outside a shop, turns a route into an obstacle course. Blind and partially sighted people cannot see a quiet machine at knee height, and older people may be startled by one approaching from behind.

Safety is usually discussed in terms of speed, and at walking pace a collision is unlikely to be fatal. But mass matters. A machine weighing twenty or thirty kilograms can injure an ankle or knock down a frail person, even at low speed, and the risk is greater when a pavement is narrow and a pedestrian steps aside without looking. And when a robot blocks a ramp or clips a buggy, responsibility is not always obvious. Was the operator at fault, the remote supervisor, the shop that sent it or the pedestrian who stepped into its path? The law on vehicles was written with cars in mind, and it does not neatly answer these questions.

Rules differ from place to place, which makes the picture harder to read. In Britain, there is no single national law for pavement robots; trials have gone ahead with the agreement of local councils, under general powers not designed for them. Some American states have passed laws recognising the machines, while some cities have restricted them to particular districts or times. The result is a patchwork in which a robot may be welcome on one side of a border and unwelcome on the other.

Design can remove some of the friction. A robot can be programmed to give way, to keep to one side, to stop at a fixed distance from a person and to announce itself with a sound rather than a screen. Geofencing can keep it out of crowded squares, markets and narrow alleys, and a charging point can be placed off the pavement rather than in the middle of it. Operators can publish the routes they use and the number of incidents, and consult disability groups before a service begins.

My own position is that the choice is not between a ban and a free-for-all. Small robots should be allowed where they replace motorised trips, and only where they meet conditions that can be checked. Those conditions should include a speed limit no higher than a brisk walk, supervision by a person who can intervene, insurance to cover injury and a duty to report collisions. Councils should be able to exclude robots from narrow streets or where residents object. The burden of proof should lie with the operator, not the pedestrian.

None of this is an argument against the technology. A small electric vehicle that carries a meal a kilometre can be a modest gain for a town. But the pavement is a shared room, and the test of any new occupant is whether it leaves less space for those who have no alternative. If robots can pass that test, they deserve a place. If they cannot, they should be confined to roads, depots and private ground.