Softbank sold 300 Pepper robots in February and that’s going to drive the artificial intelligence and app market to develop faster than most people think.
According to market research company Trendforce, while the Pepper robot (pictured) is primarily available now to the commercial market, later this year Softbank and French startup Aldebaran will sell it to the general public, fuelled by third party apps people are currently creating.
Pepper is a versatile robot and it’s claimed it can recognise peoples’ emotions, something that some people have trouble doing. The machine costs Y198,000 but coupled to that is a cloud service that costs Y24,600 per month. The inventors, Aldebaran, describe Pepper as the first social robot.
Application developers can be programmed with extra emotions and as it does so, it will react faster to its owner’s desires.
Trendforce is really bullish about this because it imagines that the arrival of Pepper the robot to the marketplace will create a market comparable to computers, to the internet and to smartphones.
The market size for service robots has already exceeded five million units and will grow by 20 percent this year, the analysts think.
Foxconn has confirmed that it is pressing ahead with its plan to replace its increasingly tricky staff with robots.
After all, robots will always work, and will not get depressed, kill themselves, or moan to western news media that they are being exploited.
CEO Terry Gou who once famously described his staff as animals, and he was the zoo keeper said that in three years, Foxconn will probably use robots and automation to complete 70 percent of its assembly line work.
Previously Gou said he hoped to one day deploy a ”robot army” at the company’s factories, as a way to offset labor costs and improve manufacturing.
This will be great news for Apple, which is Foxconn’s biggest customer, and found that its image was a little tarnished by the fact that its main supplier was installing nets to catch people who threw themselves off buildings.
Labour watchdog groups have complained that Foxconn workers have in the past faced long hours and harsh treatment from management.
Last year, Gou said that the company already had a fully automated factory in the Chinese city of Chengdu that can run 24 hours a day with the lights off.
Gou declined to say more about the factory, or what it produced, but Foxconn has been adding 30,000 industrial robots to its facilities each year, he said in June.
Gou said his company needed to adopt more automation, due to the potential for labor shortages. Young people won’t do this kind of work, and won’t enter the factories, he moaned.
A report said that home care and lawnmowing robots will be worth close to $3 billion by 2019.
ABI Research said vacuum cleaners will dominate the market and it expects over seven million robots to be in use worldwide by 2019.
Over 3.4 million robots were sold in 2014, so that represents a compound annual growth rate (CAGR) of 15.6 percent over the period to 2019.
Dan Kara, a research director at ABI, said the market is the largest, fastest growing and most dynamic sector of the consumer robotics market.
“The addressable markets for these devices are substantial, penetration levels are nowhere near saturation, and in some cases markets are virtually untouched,” he said.
He said that over the next few years we could expect to see various advances in terms of lower prices, better functionally and greater levels of performance and autonomy.
The overall market includes vacuum cleaners, floor cleaners, lawn mowers and pool cleaners.
It’s not just people
that are addicted to Youtube, according to a group of scientists from the University of Maryland.
They are teaching robots by getting them to watch online cooking videos.
The scientists say they, together with an Australian research institute, are developing robotic systems that can teach themselves.
The online cooking videos help them to learn intricate grasping and manipulation movements.
The researchers said they’ve achieved a milestone in robotic autodidactism by combining artificial intelligence, computer vision; and natural language processing.
Yiannis Alimonos, a professor of computer science said the team selected cooking videos “because everyone has done it and understands it.”
But, he continued, “cooking is complex in terms of manipulation, the steps involved and the tools you use”.
The professor said other researchers had tried to copy movements, but the team is copying the goals, letting robots decide how to combine actions. The algorithm use deep learning neural networks, which are now practical because of the vast increase in computing power and performance.
Robot operating systems (ROS) rely largely on open source software and that’s likely to continue.
That’s according to a report from ABI Research which said that using open source software, libraries, tools and the run time environment all contribute to a reduction in complexity, robustness, maintenance and speed of devleopment and deployment.
Dan Kara a director of robotics at ABI said that the complexity, difficulty and “glacial pace” of software development has stifled robotics and held up commercialisation of products.
He said that ROS is now a standard technology for robotic researchers and the next generation of engineers will be well schooled in ROS.
The move is being driven by giants like Google and Toyota, but also aided by Rethink Robotics and Universall robots.
He said: “ROS has the potential to become a foundational software platform for all manner of actuated devices, ranging from service robots, industrial manipulators, consumer systems, self drive cars and more.”
Crystal ball readers at analyst outfit Gartner have seen a future where robots and drones replace a third of all workers by 2025.
At the start of its major US conference, the Symposium/ITxpo Gartner’s research director Peter Sondergaard predicted a future where a drone may be your eyes and ears.
In five years, drones will be a standard part of operations in many industries, used in agriculture, geographical surveys and oil and gas pipeline inspections, he said.
He also predicted a rise in the a “super class” of technologies that perform a wide variety of work, both the physical and the intellectual kind, said Sondergaard.
Machines, for instance, have been grading multiple choice for years, but now they are grading essays and unstructured text.
This cognitive capability in software will extend to other areas, including financial analysis, medical diagnostics and data analytic jobs of all sorts, Sondergaard said.
Gartner predicts one in three jobs will be converted to software, robots and smart machines by 2025. The new digital businesses require less labour and machines will be make sense of data faster than humans.
Scientists at the QUT say that modelling a human eye, a rat’s brain and combining them with robots could well lead to new technology.
Dr Michael Milford, lead researcher, said: “This is a very Frankenstein type of project.” The study uses newly designed computer algorithms to let robots navigate intelligence.
Why a rat brain and a human eye? Milford said: “A rodent’s spatial memory is strong but has very poor vision while humans can easily recognise where they are because of eyesight.”
He said QUT already has software algorithms to model human eye’s and rat brains.
”We’ll plug in the two pieces of software together on a robot moving around in an environment and see what happens,” he said.
The research could also have implications for neuroscience because disease like Alzheimer’s rapidly degrade spatial navigation abilities in human brains, he said.
But it could all be a long way off. Milford said: “We’ve got all the ground work there but plugging them altogether is the massive challenge we have. I don’t know exactly how it’s going to work and that’s why it’s research.”
If the research comes to anything, we may well have a stainless steel rat scurrying round out cities.
More than 8.5 percent of Twitter’s customers are not real, but are account-holding, automated programmes.
In its most recent quarterly filing to the Securities and Exchange Commission, Twitter admitted that about 23 million, or 8.5 percent of Twitter’s monthly active users, “hold accounts that are programmatically updated “without any discernible additional user-initiated action.”
Known as Twitterbots, or simply “bots,” the programs are used for an array of different purposes ranging from the creation of revenue-generating URLs to the acquirement of instant followers for those willing to buy them.
The ease with which the bots can be created could be a problem for the social network as the the market’s confidence in Twitter is linked primarily to its viability as an advertising platform.
Twitter’s market value has fluctuated drastically since its IPO as investors pondered if the site really was worth the money. If people think that one in ten of Twitter customers is not real that they will be even more concerned.