Major government policy documents that seek to shape an economy¡¯s research direction inevitably try to identify the key fields that ministers feel are worthy of focus.

The Industrial Strategy was no exception, pinpointing four ¡°grand challenges¡± where the government wants to invite ¡°business, academia and civil society to work together to innovate and develop new technologies and industries in areas of strategic importance to our country¡±.

In addition, the strategy introduced four new ¡°sector deals¡± in areas in which the government wants to partner with industry to increase productivity and innovation, some of which overlap into similar territory.

But to what extent is the government choosing areas in which the country already excels in research and development, how much do they represent the future economic landscape, and what is university-industry collaboration currently like in each field?

³ÉÈËVRÊÓƵ

Some insight into these questions is available by looking at the data around the UK¡¯s research performance in four of the areas mentioned in the strategy: the life sciences, artificial intelligence, low-carbon technology and the automotive industry.

?

Life sciences?

Although the life sciences ¨C the subject of one of the ¡°sector deals¡± in the Industrial Strategy ¨C is a very broad area, data from the ³ÉÈËVRÊÓƵ life sciences subject ranking illustrate how the UK is already a world leader.

³ÉÈËVRÊÓƵ

For countries with at least 10 universities in the ranking, it has the highest average score for the citation impact of its research; if the cut off is five universities then only Switzerland, Sweden and the Netherlands are higher.

However, the rankings data also reveal that the aspect of life sciences research in which the UK severely lags behind competitors is the amount of money that its institutions attract from industry.

Whereas Switzerland, Sweden and the Netherlands all come in the top six for average industry income score in the life sciences ranking (counting nations with at least five institutions), the UK lags well behind despite being home to major multinational drug companies such as GSK and AstraZeneca. It suggests that a strategy of bringing academia closer to industry may be warranted.

?

Artificial intelligence

In artificial intelligence, data extracted from Elsevier¡¯s Scopus database of published research using the company's SciVal tool show that the UK currently sits fourth in the world for scholarly output in the topic, when taking into account all research published from 2012 to 2016.

³ÉÈËVRÊÓƵ

However, other countries are growing the volume of their research in this area much more rapidly: in 2016 the number of publications produced by India more than doubled to 3,700 from the year before, China produced almost 50 per cent more research and the US and Germany both grew their scholarly output by more than 10 per cent. The UK produced just over 2,500 publications in the field in 2016, a 3.8 per cent increase.

For countries producing a large amount of research in artificial intelligence (more than 10,000 publications from 2012 to 2016) though, the UK¡¯s research quality, measured by field-weighted citation impact, is second only to the US. And unlike the life sciences, its collaboration with industry appears to be better: 3.7 per cent of research in artificial intelligence in the UK over the period was co-authored with firms, behind Germany (5.3 per cent) and the US (6.8 per cent) but comparable or ahead of other tech-focussed nations such as South Korea (also 3.7 per cent).

?

Automotive engineering

Automotive engineering is where the UK currently seems to score best for its collaboration with industry: data from SciVal show that 7.7 per cent of UK research from 2012 to 2016 was co-authored with the corporate sector, ahead of the US (6.5 per cent) and only behind other car-producing powerhouses such as Germany (12.4 per cent), Japan (11.1 per cent) and France (10.3 per cent).

The quality of the UK¡¯s research in this area ¨C measured by field-weighted citation impact ¨C is also the highest of any country that produced a scholarly output higher than 2,000 over the time frame, and its annual publication rate is catching rivals such as Germany. It suggests that, in this field, the UK government is throwing more weight behind a sector that has already seen success in recent years.

However, in terms of worldwide research, automotive engineering, while growing, is nowhere near seeing the explosion in scholarship experienced in artificial intelligence and renewable energy: its global scholarly output more than doubled from about 5,400 publications in 2004 to about 13,000 in 2016, but for AI the increase was 500 per cent and for renewable energy, 900 per cent.

?

Renewable energy?

In renewable energy research, the UK again performs strongly but there is a good argument for saying that this is a rapidly expanding field in which the country does need to boost its engagement to prevent being left behind by competing nations.

³ÉÈËVRÊÓƵ

China¡¯s scholarly output from 2012 to 2016 in research related to renewable energy, sustainability and the environment was about 33,400, and it produced almost three times as much in 2016 as in 2012.

Most worryingly for other nations and, unlike in some other research areas, China also scores very highly on quality, with a field-weighted citation impact of 1.81 (world average is 1), only just behind the UK (1.84), which produced just under 7,000 publications in the period.

³ÉÈËVRÊÓƵ

Meanwhile, out of the countries with the largest research outputs in this area, South Korea and Germany lead on collaboration with industry, with more than 5 per cent of their research being co-authored with firms. The UK here is comparable with the US, with 3.6 per cent of its scholarly output between 2012 and 2016 being a collaboration with industry (the US figure was 3.4 per cent).

simon.baker@timeshighereducation.com