Monthly Archives: October 2010

Lost in space

Remember Beidou? The Chinese satellite was intended to save China from using foreign satellites for GPS systems and other types of tracking. According to Caixin, Beidou has veered off-course.

It appears that Beidou’s largest user demographic are some 10’000 Chinese fishermen in the South China Sea. Its limited use comes from the delayed release of decoding files – know as interface control document or ICD - to access the Beidou system, making it a low value proposition for application developers, not to mention the expensive terminals which don’t work with the US GPS standard.

This is too bad for a project that was intended to compete with the GPS, the European Galileo or the Russian Glonass. Beidou could end up be solely used by the government and the military or by Chinese firms since foreign companies have all been denied access to the ICD, unless they partner with a domestic enterprise. The day of open standards isn’t here yet.

Green airlines?

Will kerosene soon be replaced by biofuels to fly airplanes?

According to the Wall Street Journal, Boeing is cooperating with Air China to test a commercial jet using biofuel produced in China. The biofuel is based on jatropha, a thorny wild green shrub growing in hot climates. The good news: it doesn’t take food from people’s mouths. The bad news: in addition to costly modification to airplanes there are some worries that biofuel could freeze before reaching cruising altitude – a claim dismissed by the industry. Also, one still has to domesticate jatropha. PetroChina, foreseen as the fuel supplier, is already growing the plant in southern China.

If the technology becomes mature (potentially by 2025) and there is a concerted effort to internalize environmental externalities, one could imagine that passengers would be able to chose between flights priced on the type of fuel used.

#1 in supercomputers?

When it comes to world rankings China tends to trust the first place through sheer size rather than speed, in particular when one talks about technology. So, the recent news that it may soon be leading the ranking of supercomputers may come as a surprise.

It shouldn’t. First, China already ranks second with its Nebulae Dawning able to clock 1271000 Rmax(GFlops). Then, it is no secret that the country has embarked on a race to join the ranks of leading science powerhouses. For instance, supercomputers are used in climate modeling or genomics. Third, and more importantly, there are serious commercial prospects for high-speed computers in some industries coveted by Chinese firms, including the car and pharmaceutical industry as well as oil prospection (seismic imaging).

The next step for Chinese engineers?  To lead also the Green500 list, which measures the performance per watt.

Reforming academic journals

For those not familiar with the 21st century world of academia, the name of the game is publish or perish. In its quest to join the ranks of leading science countries, China is not spared from the lab race. There too scientists scramble to get their research results published in Chinese academic journals in order to ensure promotion and funding.

The problem? Scientists in need to publish in quantity (and in quality) are too often resorting to incremental work and plagiarism – Nature reports that 31% of the papers submitted to one campus journal contained plagiarized material. One academic estimates that only 5–10% of these campus journals (an important channel for academic publications in China) are worth saving. The government has proposed a “survival of the fittest” solution to solve this issue, i.e. “weak” journals will be terminated. But given that scientists will still need to publish, the question really is whether such a strategy has any chance to succeed. Like the phoenix, journals will be able to rise from the ashes, under a different name.

A solution that could make sense? Make sure that the Chinese journals have an English version, rigorous peer-reviewing and a ranking  mechanisms. The impact factor would take care of solving the quality issue!

After the genome, the proteome

According to Nature, BGI, a research lab in Shenzhen is positioning itself at the top of an international effort to catalogue and characterize all proteins in the human body – for those wondering there are 21,000 or so protein-coding genes in the human genome.

So far several obstacles have made this mapping exercise hard to complete: first, the cost issue (estimates run in the billion USD), second, a lack of obvious medical applications and third, even more bothering, a lack of data comparability among laboratories.

BGI, together with strong backing from the government, is set to invest substantially in the “hardware” aspect of the research. The planned new mass spectroscopy center will be a world leader in integrating the different steps in the process. But, as usual, that’s just half of the path to success. The “software” aspect (data analysis) will tell if the investment was worthwhile.