Experts at the IFR CEO Roundtable held in Chicago on April 8 debated the question of who will win the race on artificial intelligence and robotics.
Artificial intelligence (AI) and robots continue to occupy news headlines. Given the importance of the two technologies to turning around low productivity rates, which continue to hold back economic growth, most countries have prioritised the development of an AI and robotics strategy, and support for national industry champions. As Robert Atkinson, President, Information Technology and Innovation Foundation (ITIF), US, pointed out, ‘countries care about robotics because it’s a source of high paying jobs – so there’s a race to get more of these companies in each country.’ Increasingly, there is a question of which country will win the race for technology dominance. Will it be the US, with its strong track record in software development? Or will it be countries such as Germany, Japan and Korea that have a strong heritage in engineering? And what of China, whose government has put a clear emphasis on robotics as a means to improve the competitiveness of Chinese company though automation?
The IFR 2019 CEO Roundtable, held on April 8 in Chicago, assembled executives from Europe, Asia and the US to debate the issue and look more broadly about what needs to be done to deliver on the promise of AI and robotics for driving much-needed productivity increases.
Race for leadership in development and implementation of robotics
Panellists agreed that competition in AI and robotics is happening on two fronts – development and implementation. Japan and Europe are leading in robot development and supply according to IFR President and Representative Director / Chairman of the Board of Yaskawa Electric Corp., Japan, Junji Tsuda. While the Chinese robot market is growing fast, the majority of robots supplied to China come from other countries. Atkinson added that Vietnam, China and Singapore are investing heavily in robotics, but the US lags behind.
Panellists agreed that the pace of robot implementation is still relatively slow, and that the government has a role to play in accelerating adoption. Byron Clayton, CEO of Advanced Robotics for Manufacturing (ARM), US, pointed to the German model of investing in applied research institutes such as Fraunhofer that work with companies to get cutting-edge technologies on to the factory floor. He called for more investment by the US government in supporting companies in adopting robotics and AI. ‘The government needs to invest in reducing risk for the market,’ said Clayton. ‘For small-to-medium-sized companies in particular, the biggest barrier is that investment risk.’ Tsuda agreed, adding, ‘If we can apply AI to make robots easier to set up and use for SMEs, it will be a game-changer.’
Automation of service industries key to improving economic productivity and per capita income
The panellists agreed that the automation of service industries is key to improving economic productivity and per capita income. As Atkinson put it, ‘we’re in trouble if we can only automate manufacturing.’ However, Tsuda pointed out that more development is needed in robot dexterity. ‘If the application just needs a robot arm, we can do anything, but if we need a hand, that’s decades away.’ The healthcare sector was cited as a promising field for robot implementation to assist workers but, as Thomas Visti, CEO of the Danish company Mobile Industrial Robots, pointed out, ‘Two things are slowing down implementation: One is a lack of experience in working with robots, but the other is the very slow sales process – it takes years before decisions are made on tenders. China is much faster in adopting robots in the healthcare sector.’
The panellists agreed that AI will play an important role in enabling the expansion of robot implementation into new market sectors, but that the integration of AI into commercially available robots will take far longer than most people think. ‘Significant numbers of policy-makers, thought- leaders and advocates fundamentally buy into the idea that it´s only a matter of time – 10 to 15 years – before robots can do everything. That’s utter nonsense!’ said Atkinson. Even in the automotive sector, which until recently has been the largest adopter of robotics, AI is not yet widely applied. Henry Sun, Director of Strategy, Guangzhou MINO Automotive Equipment Co, China, commented that in China, ‘Most OEMs are still installing connected equipment that generates data, and you need that foundation before you can start applying AI.’ Tsuda sees huge potential for AI in digital twin applications in manufacturing. ‘AI can provide data on the real world – sounds, vibrations – that can link the cyber world to the real world and make robots more efficient and easier to use.’ All four panellists agreed that autonomous vehicles will not be in widespread use within the next five years. Sun believes that,’ We’ll see some specific use areas in geo-fenced areas, but the liability is currently too great for OEMs to take on.’
Panellists agreed that linking the cyber world to the physical world brings security issues that, according to Visti, are, ‘a huge topic – important for customers, and a learning period for us.’ Atkinson believes that government has an important role to play in reducing cyber-security risk, for example by publishing standards and then fining companies that do not adhere to them.
Skills gap slowing down pace of adoption
The group agreed that a skills gap is hindering the growth of the robotics sector, in both development and implementation. ‘It’s challenging to find highly-qualified people in AI for robotics’ said Visti. Sun agreed that the gap is in AI skills rather than robotics in China. ‘Young people are excited to be working in robotics, but it’s more difficult to find the AI experience.’ Atkinson felt the US government will need to do more to encourage students. ‘In the US, most of the students in computer science and engineering faculties are from out of state, or outside the US because there isn’t enough federal funding of state universities, so they need out-of-state income. The government also needs to do more to train manufacturing line workers to do some basic AI training so they can up-skill.’ Clayton added, ‘We need to hire for potential, not for experience because this is such a new field. Then we need to design pathways to train people’. Visti added that the private sector has a responsibility to invest in training the workforce of tomorrow. ‘In the short term, it might be hard to get the return on AI, but we need to create the opportunities – and young people are keen to move into robotics and AI’.
Faced with the question of which country will win the race for robotics and AI, Tsuda concluded that, `the AI community is open and shares new logic (programming)´. And the field of application of robotics is so broad that, as long as we can accelerate the pace of implementation, everyone will be the winner!´
Picture. © Carsten Heer