Popular Science Feature - Rethinking Efficiency

Home Authors Books Subjects Events Software Features Links Newsletter Gifts Blog Write Review What's New

Feature - Rethinking Efficiency by Andrew Price

Society is addicted to efficiency at all costs, without recognizing the hidden value of the ‘unessential’ for performance. In this article we see that robustness does more than help keep things going. Applications extend from physics and engineering to fisheries and healthcare.

The efficiency trap

Squeezing things to the limit is part of everyday life. It began in nineteenth-century industrial Europe and America, when society became seduced by profit through cheaper workforce, competition and automation. Once efficiency gained a stronghold, it became cast in stone and iron. Frederick Winslow Taylor, an American pioneer of ‘scientific management’, was particularly influential. He measured factory and office tasks with stopwatch precision and argued that productivity could be boosted if workers performed them in the ‘one best way’. Like a pervasive afterglow from the Industrial Revolution’s Big Clang, ‘Taylorism’ set worldwide standards for efficiency that remain with us today – besides, many Victorians considered it a virtue to ‘waste not, want not’. The problem is ‘Efficiency for whom?’, ‘Efficiency at what cost? and ‘Efficiency over what time?’ are questions seldom countenanced.

Narrow or ill-defined measures of efficiency influence every sphere of modern life. We seem imprisoned in a world driven by efficiency at all costs: an efficiency trap.

Driven by pressure to deliver, preventable hospital infections have become a big killer in the UK and US, and cod stocks in the North Sea are no longer there for the taking. Blindness to what can happen if things go wrong – for instance from over-optimistic catch estimates – makes matters worse: a recipe for dissatisfaction, bad decisions and, paradoxically, inefficiency. Yet yardsticks supposed to measure efficiency – the cost of achieving a certain aim (or ‘more from less’) – often hoodwink us into believing we are getting a good deal.

What’s missing is robustness, a second power, currently eclipsed by conventional views of efficiency: something in reserve – the upshot of deliberately retaining or adding unproductive baggage. Yet from this redundancy, or overkill, comes resilience, to help things keep going – even when the going gets tough.

The truth is that trading robustness for superficial saving is highly risky: the challenge lies in getting the balance right. For a chilling reminder look no further than the Challenger space shuttle. Although built to be fast and to last, it blew up in 1986 just because of a dodgy ‘O’ ring. Whether down to unfortunate engineering, or a disaster waiting to happen, adequate insurance just was not there. Worse still the 1999 Mars Climate Orbiter ended in pieces simply through failure to spot a mix-up of metric and imperial units. Being wise after the event is easy. Yet backup, through extra time spent double-checking, might have payed off. Or take the latest Discovery Shuttle Orbiter, launched on July 26th 2005. On take-off a piece of insulating foam peeled off, calling to question the robustness of this space mission too. Fortunately, it and the crew landed unscathed… though it was not plain-sailing.

Applying concepts of robustness benefits hospitals and many other systems too. The simple expedient of spending time hand-washing in hospitals, for example, can help stop 90,000 patients in US hospitals die from infection each year. Making time is not wasting time: it instils robustness, and efficiency. Seen in this light, even the traditional English Aga cooker – a stove that burns day and night – is far more than a cast-iron dinosaur that devours ten-pound notes. Besides cooking food, the Aga heats domestic water. If clothing is hoisted above it on a rack, it dries that too. Perhaps Nobel physicist Dr Dalén’s concept of efficiency through robustness, and its manifestation through the Aga, which he developed and patented back in 1922, wasn’t so far off the mark after all.

Robustness, in the form of deliberate redundancy, has some unusual applications. Take fisheries. ‘No-take’ reserves in St Lucia have, ironically, doubled the catch in a matter of years in waters where fishermen can fish. It works by the spillover effect: eggs and young fish drift from the protected area into nearby fishing areas. Conventional thinking would regard deliberately holding back as an inefficient approach to harvesting. What’s more catch quotas are often over-optimistic; some 40 to 60 per cent of the world’s fisheries are now close to exhaustion. No-take areas provides built in insurance and can help regenerate heavily exploited resources. Only by taking stock wisely can we come close to getting ‘as good as it gets’. However you look at it, robustness is a solution to all sorts of problems – including plain greed or human error.

Robustness in a nutshell

Robustness is much more than exterior sturdiness. It captures, in the words of Erica Jen at the US Santa Fe Institute, ‘our intuitive sense of one of the key determinants of long-term success or failure’.

Undeniably, robustness is a slippery concept. Unlike superficial measures of efficiency, there is no simple equation for robustness. The Santa Fe Institute website defines it in eighteen or so different, sometimes contradictory, ways. Mathematicians consider a model to be robust if is true under assumptions different from those used in building the model.

Robustness is also characteristic of systems with the capacity to heal, self-repair, self-regulate, self-assemble, and/or self-replicate. Curiously, these concepts are helpful for the understanding of yacht behaviour and performance.

British yachtswoman Ellen Macarthur’s record-breaking B&Q was fast and, on many counts, remarkably robust, too. But when it comes to ‘self-regulation’, in the form of the yacht staying on course unaided (without self-steering and with the helm lashed), Falmouth Oyster Dredgers and other old-fashioned gaff-rigged boats, with straight stems and long keels, are winners. In this sense, and in other ways, they are highly robust. Yachts also highlight the potential hazards of stripping away too much redundancy. In February 2005 Macarthur and 75 feet of pure carbon – weighing only 8 tons – arrived in Falmouth, smashing the solo record around the planet, in 71 days. So critical was weight that even freshwater was excess baggage. But relying on desalination nearly lost her the coveted prize. Despite tremendous robustness, all of a sudden the ‘B&Q system’ began to look fragile.

Theories underlying robustness are a hot area of complexity science, and leading to a new understanding of the workings of nature, technology and society. Designers (and evolution) have come up with systems that are robust against known or common disturbances – yet rare or unexpected ones can cripple them. Highly Optimized Tolerance (HOT) was the term coined by Carlson and Doyle in 1999 to describe this.

An ecosystem or species might survive a cold winter but collapse unexpectedly when a new virus or appears. Take the influenza pandemic of 1918-19. The virus responsible killed more people than World War I: twenty or even forty million. According to some, it was the most devastating epidemic in world history: a truly global disaster. Sequencing the virus’ RNA, collected from lung tissue of a dead soldier, pointed to a novel ‘H1N1 influenza A’ virus of exceptional virulence. Curiously, though, over their lifetimes, most victims must have fended off a host of other bugs and immune system assaults.

Many scientists now consider this ‘robust yet fragile’ paradox to be the most important feature of complex systems. Crucially, the required robustness is a deliberate part of the design and construction – not merely an accident of evolution or edifice of engineering. This, argue, Carlson, Doyle and others, is what creates complexity.

Rules of engagement

Much of what it takes to build, launch and keep rockets airborne hangs on super chips and many other parts. The price to pay for extremes of efficiency is heavy reliance on the system’s many different parts, and potential disaster if any goes wrong. We saw what happened to Challenger and the Mars Climate Orbiter. Other Space Shuttle missions have also failed, spectacularly: Columbia, for example, broke up on re-entry in 2003 due to foam damage to heat proofing tiles on take off.

Undeniably, ‘leanness’ can help improve efficiency; it is necessary to get a machine airborne in the first place. The price can be high, though. Superior, thicker or extra tiles might have helped Columbia, but then at what cost to other parts of the spacecraft, including its weight? Tradeoffs are inescapable, in line with the robust yet fragile principle. The truth is that there is seldom, if ever, using Robert Kanigel’s words, ‘the one best way’.

Even the military must respect this and recognize the value of robustness, for example through deliberate redundancy, or overkill. Sometimes this can be achieved very simply. The US finally tracked down Saddam Hussein; but only after seven months and a $25 million bounty on his head. Not only that, it took six hundred soldiers including cavalry, artillery, aviation, engineer and special operations forces. And that was just on 13 December 2003, the day of capture. The irony was Saddam’s spider hole, camouflaged with dirt and a carpet, was simple yet robust. It also provided safety – at least for a while.

George Bush’s second dearest wish is to capture Osama bin Laden. With a price tag on his head since 1998, Osama is used to life on the hoof. His last known hideout – several years ago – was a secret labyrinth of caves and boltholes in Afghanistan (pounded by American B-52 bombers). From there, best Intelligence estimates are that he left on foot, by donkey, or a bus into Pakistan. More certain is that the Al-Qaeda used donkeys to send messages: texting, normally efficient communication, was too dangerous given all the eavesdropping by Big Brother. High-power military hardware alone does not always ensure efficiency and a quick ‘walkover’. As the military historian Frederick Kagan reminds us: ‘Above all, the US must avoid the search for “efficiency” in military affairs. Redundancy is inherently a virtue in war. America’s leaders should intentionally design systems with overlapping capabilities, spread across the services, and should intentionally support weapons that do not directly contribute to the overarching vision of war that they are pursuing.’ But unlike spider holes or donkeys, high-tech military operations imbued with robustness through redundancy do not come cheap.

In and out of combat, all sorts of equipment and appliances are thrust upon us ostensibly to make life easier. But seen from the perspective of efficiency that values robustness, differences between ‘advanced’ and ‘primitive’ technologies may be less clear-cut. Aga cookers, for example, increasingly grace kitchens of the fortunate. But only by conventional reckonings is this technology over-costly and inefficient. Take, for comparison, a magnetic induction hob. This cooker boils water almost instantly; temperature control is as precise and immediate as the volume control on a Hi-Fi system; and unlike the Aga’s hot plates (which lose heat and efficiency when the lids are up), the magnetic hob never loses power – unless there is a power cut. Yet, in terms of robustness and simplicity, the Aga may still adorn kitchens long after the induction hob’s chips have burnt out.

New-Age infonauts

Perhaps surprisingly, robustness helps us to understand abstract concepts as well as more ‘concrete’ systems, for it gets to the heart of information itself. Take environmental assessment. Scientists can gather information using low or high tech approaches, which capture very different things: they vary in efficiency, but how much so depends on reliability of the equipment and observer, on accuracy, precision, generality and so on that is required, acceptable and – indeed – possible.

Robustness provides a helpful way of considering this. Our work in the Red Sea during the 1980s helps explain why. This is a region whose reefs dwarf even the most ambitious structures created by modern technology, and are justly world famous.

Like the gutras tied around our heads to block the sun, the technology for our mission was simplicity itself. But the noon heat still reached parts that Heineken could not, especially as alcohol was forbidden. The Suzuki and Land Rover stopped at the first of over a thousand sites we set out to survey. Besides vehicles, we had little more than a mask, snorkel and waterproof paper as we ‘eye-balled’ the abundance and condition of coral reefs and other systems along Saudi Arabia’s 1,500-km Red Sea coastline, rating them on a simple score of 1 to 6.

This can hardly have been accurate: a score of ‘1’ is for, say, bird (or turtle) numbers anywhere between one and nine; ‘2’ covers the range ten to ninety-nine; ‘3’ is any number from one hundred to nine hundred and ninety nine, and so on. The approach seems very imprecise, and the truth is it is. But this ‘slop’ in the scale is also precisely why it did not matter when one ecologist estimated a bird population as 500, whereas another’s score came closer to the accurate value of 670 – painstakingly counted by a real ornithologist. Either way, with rapid assessment – providing low-resolution data is acceptable – it doesn’t matter; the right value is still ‘3’ because the (logarithmic) scoring system is robust against differences in observer experience.

More than a decade later, in the mid-1990s, I was one of several marine scientists at work in Chagos. Best known for the US military base on Diego Garcia, and the unlawful eviction of its islanders more than 30 years ago, this British Indian Ocean Territory sits in the most isolated patch of the Indian Ocean. The question was: does Chagos, as one of the remotest and least frequented parts of the world, remain pristine and contaminant free? Crude eye-balling was no use here. Instead, we collected samples of coral and lagoon sand for ‘environmental forensics’ using gas chromatography: GC-FID and GC-MS and other precision instrumentation.

Only this could yield information on petroleum hydrocarbons, PCBs or pesticides in the trace concentrations at which they might be present. The only trouble is cutting-edge instrumentation demands surgical cleanliness: scientists must be on guard not to contaminate the samples, whether from sun-screen on their bodies or from outboard motor fuel. Otherwise, false alarm bells may ring unnecessarily when results come through from the lab. Human skill matters greatly– during sample collection and analysis. You could say going high tech means lower robustness against differences in scientist experience. What a contrast to rapid environmental assessment.

Data from rapid assessment might seem rough and ready. But that is nothing compared to the strange world of expert systems and fuzzy logic. Its applications now extend from washing machines, PalmTop computers and self-focusing cameras to coastal management and mine detection on the seabed. Paradoxically, uncertainty and ignorance are here seen as critical and dealt with head-on (rather than ignored) using a mathematical ‘fuzzy’ score. Fuzzy it may be yet, like rapid environmental assessment, the approach seems robust.

In his book Dreaming in Code, Scott Rosenberg reveals an even bigger paradox. At the end of the day most computer programmers entrust their meeting notes to notebooks not bytes; foolscap in their eyes is more foolproof. This resonates with the growing consumer demands for traditional products. So-called efficient goods and services are increasingly thrust upon us, but ought we not wonder if their assumed efficiency is rooted in something sensible in the first place? We welcome handheld computers to remind us where to be at 2.30 pm tomorrow, yet if they lack reliability then we might as well trade microchip speed for Filofax robustness.

Robustness even helps us understand how our brains process information. Maltese psychologist Edward de Bono coined the term ‘lateral thinking’ in 1967. Instead of looking ‘harder in the same direction’ for solutions to problems, you look at an issue from side to side, from unexpected angles. It takes time. On the face of it, this cerebral detour might seem an inefficient way of problem-solving. Often, though, trying to get there directly simply does not work. So ‘unessential’ pathways of thought can be efficient after all.

However, you look at it robustness is little short of a universal solution to wide-ranging problems, including human error and plain greed. It helps explain what Spanish conquistadors had (through gunpowder and horses) that the Incas lacked, which in 1533 made their civilization tip and suddenly become fragile; why extra time spent by patients recovering in hospitals can improve long-term healthcare efficiency … and much more besides. Yet robustness is far more than insurance. Hit the right level and the payback is performance that lasts, the quest of every efficiency-seeker.

Andrew Price is a Reader at the University of Warwick and an international marine/environmental consultant. He has expanded the article's contents into a book, "SLOW-TECH: Manifesto For An Over-Wound World" , which is due to be published in Autumn 2008 (Atlantic Books).

DISCLAIMERS

This site has no connection with Popular Science magazine or other sites and publications with a similar name.

Much of the content of this site is written by popular science writers or friends of popular science writers. Inevitably many of the reviews in such a small community are written by or about someone we know. We always aim to be impartial in our reviews, but there is a connection which we need make clear, as there is no intention to deceive. The content of any review or article is solely the opinion of the author and should not be read or understood on any other basis. The site exists to promote popular science writing and popular science authors and for this reason should be considered promotional material, just as the editorial reviews in an online bookshop or the blurb on the back of a book should be considered promotional.

The website should not be eaten or used where it can come into contact with water.

Disagree with our review? Want to comment on a feature? Contact us at info@ popularscience.co.uk - have your say!

Part of the site