The government's target of a 60 per cent reduction in carbon emissions by 2050, announced in March, is optimistic by any standards. Nor does its disjointed approach to energy research help: specialists on, say, wind power and plastics recycling share the same ultimate goals, but are barely aware of each other's existence. That's why the rationale behind the Energy Technologies Institute (ETI), announced by Gordon Brown in last year's budget, is sound. The ETI will provide a hub for all Britain's non-nuclear low-carbon energy research, whether that means improving the efficiency of fossil-fuel use or doing away with it entirely.
The ETI is a public-private partnership: the government will match industrial funding pound for pound. When fully operational (in 2008, with luck), it will accrue £100m a year, half provided by ten commercial partners that so far include Shell, BP, E.On, EDF and Rolls-Royce. These companies will largely determine the research objectives, something which may worry environmentalists—but needn't. Even cynics can't still doubt the real commitment of energy companies to alternative sources, which after all makes commercial sense.
It is hard to imagine public money alone funding a project like the one supported by BP on the Scottish east coast, where conversion of North sea gas to hydrogen will supply enough zero-carbon fuel to power a quarter of a million houses. The process makes carbon dioxide too—but this will be pumped back into the ground, specifically into the porous rock of the ageing Miller Field oil reservoir, which not only stores it out of harm's way but helps to squeeze out more oil from a field that would otherwise have soon been abandoned. The technical challenges involved in a scheme like this are just the kind of thing the ETI will address. The institute doesn't yet have a home, but the tenders are in. My hunch is that we will see another challenge to the hegemony of the London-Oxbridge triangle from the burgeoning far north.
Research funding goes metric
Less widely noted in Brown's 2006 budget were dramatic plans to change the way university research departments are funded. No more expensive and time-consuming panels of reviewers deciding on a five-point rating of quality, as in the existing research assessment exercises (RAEs). By 2008, the "firm presumption" is that "the system for assessing research quality and allocating funding will be mainly metrics-based." In other words, it'll be done by a mathematical formula.
What sounds worse is that the metric will be based not on the quality of a department's work but on how much funding it already draws in. But the proposal is not as venal as it sounds: its reasoning is that the funding determined by RAEs essentially mirrors that from specific (peer-reviewed) grant applications to the various research councils. In other words, the same job is being done twice. Interestingly, funding from other sources (primarily industry) also tracks the RAE figures. On this basis, an input-based metric makes sense.
But the whole issue of metrics for calculating research quality has recently become a big deal. Researchers rightly say that merely counting up papers published in high-profile journals is deeply flawed; not least because it skews publication behaviour. In any event, getting published in Nature or Science is no guarantee of quality or significance. A number of more ingenious metrics have been proposed for measuring the real impact of a paper or an individual's research oeuvre, but although such exercises are always good value for making "greatest hits" lists, no single formula can ever encapsulate all the criteria for importance, creativity or fecundity. Peer review has flaws too, not least its subjectivity. But there's still no convincing case that assessment can be automated.
Robotic behaviour
South Korea has just announced the creation of a "robot ethics charter," which will, among other things, compel robots to observe something like the famous three laws of robotics (robots must not harm humans, must obey orders and must preserve themselves, in that order of priority) devised by Isaac Asimov in his 1950 book I, Robot. Science fiction fans will know, however, that Asimov's point was to explore how these laws contain loopholes and logical contradictions. Moreover, the laws were intended to function at the cognitive level, which is a bit futuristic for today's assembly-line machines. So it could seem silly to instigate such "robot ethics"—unless this is an attempt to imply that South Korea's robotics programme is more advanced than it is. But the Korean charter's aims are primarily short-term: placing constraints on the way robots are used, ensuring machines are traceable and so on. Still, the notion of robot ethics raises interesting issues, as there is considerable military interest in robotics, not just for surveillance but for combat—the US already has an unmanned airplane equipped with a laser, called Armed Predator. So South Korea's decision appears to be an expression of intent to which it is hard to imagine the US signing up to.
