Late October spelled bad news for Pepper, the quintessential and iconic face of Softbank Robotics. According to a Nikkei Tech survey, only 15% of companies planned to renew their three-year contract with the robot provider that began in October 2015. The contract involved deploying Pepper in corporate settings, including banks and offices, for purposes of entertainment, hospitality and so on.
The fact that few wished to renew is a humbling sign to the Softbank behemoth, which has gobbled up robot hopefuls like Aldebaran and invested in well-known manufacturers like Boston Dynamics. In the case of Pepper, the lack of a clearly defined use case has led to questions about its long-term viability as anything other than an interesting prop. When 23 out of 27 companies surveyed do not plan to renew, it suggests Pepper may not be long for the commercial market.
Pepper’s struggles are likely part of the reason for Softbank’s expansion into more practical robotics applications with its new robot Whiz, an autonomous floor cleaner. The Whiz robot will go on sale in Japan in February 2019 under a robotics-as-a-service (RaaS) model for $222 per month. The Whiz robot uses self-driving software and sensors from Brain Corp., a San Diego-based startup that raised a $114 million funding round, led by the Softbank Vision Fund, in mid-2017.
Softbank, which announced the Whiz robot at a news conference today in Tokyo, said it comes with a handle that a human uses to “teach” it the layout of the space. Once you teach it the space, the Whiz robot can perform the cleaning autonomously. The Whiz robot features a laser range finder, 3D camera, collision sensor and a battery that lasts three hours.
Challenging times for robotics companies
There are multiple reasons companies are not renewing their Pepper contracts. One executive said, “the boom has gone.” Whatever novelty Pepper may have had three years ago, the robot is costly, takes up space in the lobby, and its functions can be easily subsumed by cheap human labor or other technologies like web services. The lack of price competitiveness coupled with no clear mission suggests Pepper’s future as a viable commercial product is moot.
This has been a challenging 12 months for the robotics industry. On the consumer side, Jibo and Mayfield Robotics folded due to high cost, lack of identifiable market, and inability to compete or distinguish themselves from the excellent range of cost-effective smart home devices offered by Amazon, Google and others.
On the industrial side, Rethink Robotics’ pioneering vision was not enough to mask a flawed technical product and the company was not able to focus on its niche use case as a research tool. For sUAS, Airware tried to be a comprehensive operating system for the industry and ended up burning through tens of millions of venture capital.
Many investors have called the industry’s bluff and suggested the robotics opportunity is overvalued. I disagree. Robotics is being adopted at an accelerating pace. At the level of industrial robotics, Chinese demand and the re-emphasis on manufacturing in the developed West is leading to significant growth. There is reason to believe this will only become more so with the issues of acute labor shortage and aging populations.
Mobile robotics is experiencing even faster growth, with ABI Research predicting the warehousing market for commercial robots to grow from USD $2.5 billion in 2018 to USD $ 30 billion in 2027, with most value stemming from materials handling.
Focus on a solution, not a product
The issue is the emphasis of the robot as a product, as opposed to a solution. No industrial robot makes sense without a much larger hardware and software architecture behind it, and no individual arm makes sense outside of the context of a clear use case where the robot deployment clearly provides productivity gains. Given the state of current robotics technology, companies are far more likely to succeed by building their system around a particular use case.
This has been effective for iRobot and Kiva Systems (now Amazon Robotics) for instance. In the case of Softbank, the strategy appears to have been to develop highly sophisticated systems and wait for the market to take them on. This has not worked. There simply is no possibility of a general-purpose robot that can be fitted to multiple tasks across multiple industries.
The closest we have to this is Sarcos Robotics, a company that is successfully marketing two highly adaptable robot systems (a mobile surveillance system and a tele-operated mobile arm platform) that work on a RaaS model. In these cases, the platforms have enjoyed adoption in niche use cases, such as material handling on nuclear decommissioning sites, and are likely to succeed further due to Sarcos’ genesis in the US military. But this sort of model, with highly flexible systems with multiple use cases, takes decades of research to incubate. When the market is mining or legacy nuclear infrastructure, that is one thing, but when its B2C, as it was with Jibo, the chances of success are slim.
Pepper’s struggles are unlikely to deter Softbank from the robotics market at large. Its investment in proprietary OS developer and navigation provider Brain Corp makes Softbank well placed to be a key player in mobile robotics going forward, especially now with the Whiz robot. As Brain Corp is OEM-agnostic, it has a wider appeal than investments in Boston Dynamics, Pepper or robot pizza maker Zume.
The companies that help galvanize robotics adoption will not do so by investing in novel hardware technology. There needs to be a stronger focus on enabling technologies that make robotics possible, including better drives, cloud robotics platforms, IoT microlocation sensors and alternative operating systems to ROS.
About the Author
Rian Whitton is a research analyst for ABI Research, a market intelligence company focused on the most transformative technologies and their impact across industrial, commercial and consumer markets. As part of the Strategic Technologies research team, Rian provides an analysis for Robotics, Automation, Intelligent Systems, Artificial Intelligence and Machine Learning. He has also written actively on the commercial application of unmanned aerial vehicles. He has been a speaker at the Robotics Summit & Expo, produced by The Robot Report, and regularly contributes to media outlets like Bloomberg, Thompson Reuters, ZDnet, Recode and the Financial Times.
Prior to ABI Research, Rian graduated in 2017 with a Master’s degree in Science & Security from King’s College London, researching the intersection of technology and defense. He has chaired panels on and spoken at academic events about the intersection of technological innovation and defense, particularly in relation to AI and Robotics.
Filed Under: AI • machine learning, The Robot Report, Robotics • robotic grippers • end effectors
William K. says
Any device that does not deliver a function with value exceeding the cost is a toy, not a tool. Certainly there is a place for toys but it is not the same place. So the cute robot that is fun to see and entertaining is a toy, if it is not able to provide a benefit at least equal to it’s cost. Marketing to industry almost universally demands that the product deliver a real value exceeding the products total cost. That should not come as a new insight to anybody. So a robotic product, like any other, must be designed to provide an adequate benefit (Use Case) so that the potential buyer will be convinced. “Cute” only works for toys,and even then there needs to be some realizable benefit, such as fun.