Registered users can unlock up to five pieces of premium content each month.
Lack of Skills and Direction |
NEWS |
In an event which pivoted around leveraging new technology to improve economic growth, the most evident theme across the talks was a sense of frustration among economists, policymakers and industry leaders about the lack of sufficiently skilled Science, Technology, Engineering and Mathematics (STEM) graduates entering the workforce.
Mark Preston, the CEO and founder of StreetDrone, an autonomous vehicle company, argued the UK needed 1.8 million new engineering graduates by 2025, and that the demand for new engineers translates in EU€27 billion for the British economy each year.
On current form, the UK is failing in its ability to allocate graduates to these industries, which are in serious demand of skilled workers. The Higher Education Statistics Agency (HESA) found that in academic year 2015-16, the number of first-degree students in computer science, engineering & technology, and physical sciences were all lower than the levels for medicine, biological sciences, social studies, business & administrative studies, and creative arts & design. These patterns were in large part replicated for postgraduate degrees.
This problem is not merely a misallocation of graduate courses, but a strategic concern for the British education system, as illustrated in an early 2016 report by the OECD. The report found that, while the retiring generation (55-64) of English citizens had similar levels of numerical skill to their foreign counterparts, in the 16-34 age bracket, England ranked 21st out 23 developed countries, with 48% of the demographic lacking basic numeric skills, in contrast to the OECD average of 29.8%.
Professor Phillip Bond echoed similar concerns during his discussion regarding the need to align AI and Robotics to future economic growth and rises in productivity. A further issue was the failure of successive governments to align research with industrial strategy. Despite amounting to 11% of GDP, the UK manufacturing sector was related to some 70% of the government-funded research, suggesting a lack of research investment on the part of the service economy.
These problems are not isolated in the UK. The United States also preformed below the average for OECD countries in the same report, and the openings in the resurgent manufacturing sector are not being met with a surge in employment.
Autonomous Vehicles Becoming Accessible |
IMPACT |
There were a number of strategic technologies being discussed at length, with an emphasis on chatbots, neuromorphic networks and consumer robotics. However, perhaps the most promise can be attributed to autonomous vehicles. Mark Preston, the CEO of Oxford-based autonomous vehicle (AV) startup StreetDrone, talked about how the private industry is finally beginning to leverage the lowering cost and increased accessibility of self-driving and electric vehicle technology. The future of vehicular transport, as StreetDrone sees it, is autonomous, connected and electric.
While there are over 30 major companies researching and investing in driverless technology, a disproportionate amount of the dividends are going to a small selection of large manufacturers. StreetDrone’s aim is to make self-driving technology, research, and testing more available to a wider audience. For this, they have formed a strategic partnership with French automobile giant Renault, and have modified their tiny electric Twizy vehicle with self-driving capability. Modifications include the addition of sensors, stereo cameras, LiDAR, and a laser safety curtain.
This platform is designed to provide a viable test-vehicle and self-driving hardware, something out of the reach of most small-to-medium size software developers. StreetDrone have incorporated their own operating system with which software developers can experiment, with the hope they will publish their developments on a bespoke open-source database. This offers the ability for small vendors to access innovation with unprecedented ease.
There is increasing momentum behind the push for autonomous and electric vehicles. The costs for EV’s is decreasing rapidly, while the switch from personal ownership to autonomous carriers offers the chance for far fewer vehicles on the road, lowering emissions while alleviating traffic and congestion.
StreetDrone has initiated a useful platform that will accelerate innovation in the market and make testing and research more accessible to a greater range of actors.
IBM Insight Into Neuromorphic Networking and Quantum Computing |
COMMENTARY |
Representing IBM at the conference, Peter Waggett gave a talk about two technological trends that are expected to make a major impact on machine-learning and artificial intelligence, neuromorphic networks and quantum computing.
Waggett’s talk discussed IBM’s advances in neuromorphic computing in 2016, particularly the research team’s development of the 2016 TrueNorth neurosynaptic chip. This latest product was further developed from the 2014 SyNAPSE chip. A neurosynaptic chip operates differently to a traditional computer; beyond being able to pack more transistors in (5.4 billion), SyNAPSE had 256 million synapses and 1 million neurons. Despite the increased sophistication and power, it consumes far less energy than traditional computers. The ability of neurosynaptic chips to address sensory and pattern recognition opens a new frontier in deep learning and machine-learning techniques, which will inevitably spur innovation in the AI and Robotics space. Waggett discussed future innovation with synaptic plasticity, by which this sub-group of emerging computers could develop online learning. In a demonstration of recognition software, Waggett showcased the neuromorphic computer’s ability to understand nuance and complexity when differentiating between pedestrians, cyclists and those on scooters.
In terms of quantum computing, IBM have been one of the early movers, showcasing the experimental IBM Q. The makeshift device is a major step in trying to commercialize quantum computers for a variety of industry specific tasks. Quantum computers are best applied to problems where patterns cannot be easily deduced and there are more potential possibilities than even the most powerful supercomputers can process. Areas where this would have an effect would include chemistry and electro-magnetics.
Coda |
COMMENTARY |
Overall, the AI and Robotics Main Event gave a mixed picture on the state of the sector. They were manifold individual cases of optimism related to startups like StreetDrone, while IBM’s explanation of the neuromorphic and quantum computing technologies suggests rapid progress. On the other hand, there was a general concern in relation to the challenge the UK has in leveraging new technologies to create an effective industrial strategy, and interest in how education might be improved to address what is expected to be a growing shortage of skills in the labor force.