Robotics & Geopolitics: Global Supercomputer Race Creates AI Opportunity

Interior of Byron's concept vehicle, which will include autonomous features. Source: Byton Ltd.

June 15, 2018      

Countries and companies are rushing to score goals in proving their leadership in artificial intelligence and robotics. The AI opportunity is no doubt significant, but they will have to determine their priorities beyond bragging rights.

The nations and institutions supporting research must connect their initiatives to practical, profitable products for true competitiveness. This could be something as humble as a weeding robot or as flashy as a powerful supercomputer for new AI insights.

Robotics Business Review has partnered with Abishur Prakash at Center for Innovating the Future to provide its readers with cutting-edge insights into recent developments in international robotics, AI, and unmanned systems. Are you ready to be updated?

Switzerland’s new export is robots, not tax havens

Robotics development: EcoRobotix, a Swiss robotics firm, has been developing a robotic weeder that uses cameras and AI to spot weeds and kill them without affecting other crops.

EcoRobotics robotic weed killer

A robotic weed killer. Source: EcoRobotics

The solar-powered robot uses two arms to identify weeds and dispense herbicide. Because of its accuracy, the robot uses 20 times less herbicide than traditional weed-killing machines, claims EcoRobotix, which raised $10.6 million last month.

Geopolitical significance: Tax havens used to be Switzerland’s claim to fame, but in the future, it could be robotics. The weed-killing robot isn’t the first Swiss robot to make headlines.

Switzerland is the “Silicon Valley of robotics,” Chris Andersonn, CEO of 3D Robotics, told Forbes. Swiss researchers are building everything from Tribot, a robot that can perform versatile search-and-rescue missions, to Ultra-fast, a robot arm that can quickly and accurately catch thrown objects.

The Swiss Federal Institute of Technology Lausanne (EPFL) is also working on “robot nerves” that can be used in smart clothes, as well as a robot nose to help in disaster management.

Also, Switzerland isn’t part of the European Union (EU), so it isn’t as susceptible to the economic and political “ebbs and flows” of the EU as other countries are. This makes Switzerland a unique and stable market in Western Europe.

However, Switzerland does face certain challenges as robots emerge. The strong Swiss franc and high wages are pushing Swiss companies to turn to automation. One Swiss fragrance maker invested $60 million over three years into automation to increase its output by a third (no humans were added).

Moreover, the chief executive of Ricola, the maker of a Swiss medicinal candy, said the company is trying to automate operations from its back office through production. As Switzerland exports robots to the world, its economy will grow. But as its economy grows, the country itself could depend more on robots.

China, U.S. accelerate in self-driving race

Robotics development: Byton, an electric car startup from China, has unveiled a new self-driving sedan concept. The car can drive itself in “most conditions” and will make use of lidar technology (which Tesla has forgone).

Byton is working with Baidu for self-driving technologies and said it hopes to have its sedan enter production by 2021. Byton is also working on an SUV that will have autonomous capabilities (although less than the sedan) and will enter production in 2019.

According to Byton, both the sedan and SUV will be 40% cheaper than a Tesla Model X.

Geopolitical significance: In a short span of time, self-driving cars have become a new battlefront in the rivalry between the U.S. and China. These vehicles represent the next phase for the automotive industry and an AI opportunity. The sale of these vehicles around the world might bring unprecedented returns to companies.

Byton concept SUV shows China pursuing AI opportunity

Byton says its SUV will include self-driving features and be cheaper than other models. Source: Byton

However, China is building its self-driving capabilities in a way that could hurt foreign technology companies and automakers. In 2016, China released its roadmap for self-driving cars. It wants full self-driving car by 2021, and Byton appears to be operating with this timeline.

Between 2026 and 2030, China wants every vehicle to have some autonomous capabilities. Just this year alone, Baidu, Tencent, and Didi have received approval to test self-driving cars. (Didi will be testing them in California.)

As China plans to deploy 30 million self driving cars within a decade, it is explicitly using only Chinese technology.

Companies like Horizon Robotics are being tasked with creating chips for self-driving cars, not Intel or Nvidia. The global market for chips in self-driving cars could reach $5 billion by 2021.

The most important thing is that China isn’t allowing foreign companies to map Chinese roads. If foreign automakers and suppliers don’t have accurate maps of China, then their self-driving cars won’t be able to operate well.

U.S. companies such as Tesla, Uber, Google/Waymo, General Motors, and Ford are all making progress in self-driving cars, but they may be operating based on the assumption that they will can sell the way have in the past. This won’t be possible.

China wants its self-driving cars to lead the world, not America’s. With price, technology, and mapping, it is taking the steps to make this a reality. 

IBM supercomputer

An IBM supercomputer. Source: Pixabay

AI opportunity for U.S. as Summit surpasses Asian supercomputer

Robotics development: The U..S. has launched the world’s fastest supercomputer, called “Summit.” Developed by IBM and Nvidia, Summit has 200 petaflops of processing power (200,000 trillion calculations per second).

Summit surpassed China’s TaihuLight supercomputer, which had 93 petaflops of processing power.

Speaking about the AI opportunity at the Summit unveiling, the U.S. Secretary of Energy Rick Perry said, “We know we’re in a competition, and it matters who gets there first.”

Geopolitical significance: Supercomputers are fast becoming a measure of the computing power of a nation. While the U.S. may have the world’s fastest supercomputer, it no longer has the most supercomputers — 143 to China’s 202.

Other countries are also involved in this competition for processing prowess and the corresponding AI opportunity. In 2017, Japan was working on a 130-petaflops supercomputer. Before Summit, it would have been the fastest in the world.

Last year, Russia announced a new supercomputer that can process 55 trillion calculations a second. In 2018, India showed off the world’s fourth fastest supercomputer in the world for weather and climate research. By 2022, South Korea wants to have its own supercomputer, capable of 1 quadrillion calculations per second.

The big question for the countries deploying supercomputers and the companies building them is how the supercomputers will be used? Traditionally, supercomputers have been more about chest pumping than application. But now, countries are taking advantage of AI opportunity to apply them in different ways.

For example, in 2016, Russia unveiled a supercomputer to help predict wars. And later this year, Japan will launch a supercomputer solely to research nuclear fusion.

As nations deploy supercomputers to solve specific problems or research important areas, it may push other nations to buy or develop supercomputing technology. This creates a huge AI opportunity for businesses to supply supercomputing technology over the cloud.

For example, instead of building a supercomputer, the United Arab Emirates (UAE) is using IBM Watson’s Cognitiv platform.

While the U.S. has regained its crown as having the world’s fastest supercomputer, it may not be for long. Word is that China is prepping a supercomputer, slated for 2020, that will measure in exaflops, not petaflops. It will be able to process one quintillion calculations per second, or 1 followed by 18 zeros (1,000 000 000 000 000 000).