Retail and Web services giant Amazon is said to be working on a new breed of domestic robots, in a project codenamed ‘Vesta’ – after the Roman goddess of the home.
While details of the project are unclear, the most likely scenario would be an evolution of its Alexa personal assistant AI, coupled with computer vision, mobility, and connected home functions.
Many low-cost domestic robots already include functions such as home security, diary reminders, and chat, which in Amazon’s case could be linked with text-to-speech and infotainment functions, via its vast Kindle, music, and app libraries.
According to Bloomberg, the project began a few years ago, but there has been a recent spike in hiring activity in areas such as robotics software engineering and sensors, which suggest that formal announcements may be imminent.
Rumours suggest robots could be made available to Amazon staff this year, and released to the public next year.
Former Apple executive Max Paley is leading the work on computer vision, reports Bloomberg, and Amazon has also hired specialised mechanical engineers from the robotics industry.
• In related news, Amazon has teased the imminent debut of the Alexa-powered Fire Cube TV set-top box, suggesting that Alexa will be core to most, if not all, Amazon consumer hardware in the near future. Amazon is also releasing child-friendly updates for its Echo smart speakers, according to this report, suggesting that the children’s market is an important one for the company. Story-telling apps for robots would be on obvious extension.
Internet of Business says
Internally in its warehouses, Amazon is one of the most highly automated companies in the world, via its Amazon Robotics subsidiary, using technology acquired from Kiva Systems in 2012.
However, the domestic and/or humanoid robotics market is a different matter: it has long been an uneasy mix of promise, fiction, and disappointment. While humanoid robots such as NAO and Pepper – originally developed by France’s Aldebaran Robotics, now owned by SoftBank – have proved popular in Japan, they are far too expensive for most home budgets.
NAO, Pepper, et al, are impressive engineering feats, but there are too few apps for them, and – although they can be connected to platforms such as IBM’s Watson in the cloud – much of their onboard speech/recognition functionality revolves around trigger phrases and pre-programmed actions, rather than true natural language conversation. As a result, their initial promise and capacity to entertain hides limited functionality and, often, disappointment.
The lack of natural language AI within most humanoid robots has been a drag on their wider uptake, as – outside of specialist applications in education and retail, or research and development in robotics labs – it has been unclear what their real purpose might be.
However, into the breach in recent years has come a generation of low-cost domestic robots, which are little more than smart toys in some cases, or home hubs on wheels in others – many of them produced by electronics companies such as LG and Bosch, and startups such as Blue Frog (whose Buddy robot is pictured above).
While other manufacturers, such as Sony, have brought out domestic robots over the years, these have again been expensive. The Aibo dog had its fans as a pricey lifestyle accessory in the Noughties, while Sony’s impressive QRIO humanoid – which predated Aldebaran’s NAO – was perhaps too early to market, and was not developed further as a commercial project.
As a result, a major opportunity still exists for a company to produce robots that are accepted en masse into the home – which demands that they have a clear purpose, and can perform tasks that a smartphone, smart speaker, tablet, or Echo/Dot device are unable to. Apps and content will be the critical factors – beyond the world of single-task devices, such as robot vacuum cleaners, of course.
Amazon’s advantage over others in the Western market is that it can regard its hardware – such as the Echo, Dot, Kindle, and Fire ranges – as loss leaders for its subscription and retail services, drawing customers into an ever deeper relationship with the company. This could keep prices low and overcome the core problem with home robots, which is that they are too expensive for their limited functionality and intelligence.
This is why it seems inevitable that whatever emerges from the Vesta programme will be linked to Amazon’s retail platform and other services, such as Prime. Meanwhile, for developers, the hidden promise would be integration with Amazon Web Services and related technologies.
Even so, the commercial viability of the domestic robot market remains uncertain, and – to a large degree – a hostage to 100 years of science fiction stories. Countless humanoid and domestic robots exist throughout the world, but their purpose is largely unclear, apart from the popular trend in Japan for cute companion robots, such as Toyota’s Tomotaka Takahashi-designed Kirobo Mini (video link).
When Google bought Boston Dynamics in 2013, it appeared to herald the coming of the robot age as envisaged by a century of lore: intelligent robots linked to all of Google’s services. But within three years, Google had decided that it would take years of ongoing investment to see a return – if any – on the deal. By 2017, it had abandoned its ambitions for the company and sold it on to SoftBank.
Before the Japanese company stepped in, Amazon had been suggested as a possible buyer.
- Read more: China’s plans for global robotics dominance gather pace
- Read more: Tesla too reliant on robots for manufacturing Model 3, says Musk
- Read more: DHL US trials robots, AI, AR & crowdsourcing to beat Amazon
- Read more: South Korea most automated nation on earth, says report. The UK? Going nowhere