Sony used to be regarded as one of the very best technology companies in the world, but in the past few years the electronics giant has claimed that it’s had to deal with a few setbacks. Its mobile phone business is not as successful as it would like it to be, and the market it was hoping would lift its fortunes – sensors – is also said to be going through a less profitable phase.
It’s difficult to know what mind games these large, multinational corporations like Sony are playing, because even as it gives the impression of having to deal with struggles and challenges, its profits from smartphones are up 133 per cent, according to a report CNBC.com. Up 133 per cent? There’s probably plenty of companies who’d welcome having to struggle with 133 per cent increases in profits. Unless Sony staff are having to take all the money in rather large and heavy coins, one wonders what the struggle is.
However, Sony’s sensor business is reportedly not doing as well, with lower sales and fewer orders. Sony had previously said it would concentrate more on the sensor market, suggesting that it believed the technology was its way out of the current “struggle”.
Perhaps Sony’s lower revenues are because sensors are becoming cheaper to buy, partly as a result of them being mass manufactured for smartphones, benefiting from economies of scale. The global success of smartphones is a modern phenomenon, with anywhere between 2.5 billion and 6 billion units currently in circulation, depending on whose calculations you believe. Whereas previously a sophisticated sensor that can do something clever, like recognise you from your fingerprint, may have cost a lot of money, now, with Apple incorporating the technology into its iPhones, hundreds of millions of the things are being manufactured. At that scale of manufacturing, the unit cost is much lower than it would be if the sensor only featured in Sony phones, which it did initially.
Apple may have reached understandings with Sony that went largely unnoticed, except by iPhone users who may have coveted the Sony technology and hoped Apple would adopt it. The patent for that type of technology was said to have been applied for by Sony several years ago. In any case, Sony is a big supplier of image sensors to Apple, and continues to sell sensors in colossal quantities to many smartphone companies, even if lower unit prices may be negatively affecting the company’s profits.
Among the many agreeable outcomes of cheaper sensors is that researchers and designers of new, sensor-based technologies can now incorporate sensors into their prototypes without worrying too much about the costs.
Towards a world where there are no checkout tills
Here, in this interview, Dipak Raval, commercial director at the company, speaks exclusively to Robotics and Automation News about one of the company’s latest inventions – a smart shopping trolley. The clever thing about the trolley is that the sensors in the wheel collect power from the rotation of the wheel and could use that power for other parts of the trolley, perhaps a screen display on the handlebar, with a map showing where bargains are located, or to the shortest queues. It’s more likely to communicate to the customer’s smartphone.
The smart trolleys could use Bluetooth, near-field communication (NFC) or some other connectivity system to communicate with a central computer, which could track their movements – perhaps relaying the data to a smartphone or tablet app, which would help shop managers organise their stores more effectively, saving a lot of money on lost trolleys.
Cambridge Consultants says the technology “could spell the end of queues, as it can be used to alert staff when customers are nearing checkouts, allowing them to automatically allocate assistants when they are needed most. And it could drastically cut the multi-million-pound annual cost of lost shopping trolleys as businesses could set alerts when they were being removed from their premises”.
Raval says the cleverest thing about the smart trolley the company is demonstrating is not the hardware, but the software.
“The key thing about the smart trolley example we are showing is that it uses standard commodity low-cost, low-power sensors,” says Raval. “There is nothing special about the sensors, they are typical of the ones you will find in today’s consumer products and smartphones.
“The clever aspect of this example is that we can take really noisy sensor data – from cheap sensors – and then add our clever positioning algorithms that are able to present a very accurate location. The clever bits are in the mathematics and algorithms, not the sensors.”
And the other cleverest thing about the smart trolley is of course the wheel. Cambridge Consultants could be said to have reinvented the wheel, or at least made more of the x,000-year-old invention than most. As Raval explains, the sensors capture energy from the rotation of the wheels.
“We are able to put low-cost sensors in a small wheel and are able to show that the rotation of the wheel is sufficient to power all the electronics,” says Raval.
“This smart wheel can easily be substituted for a normal wheel to upgrade the trolley to be ‘location aware’. Even with this, the retailer can see where all the shopping trolleys are within the store, and see if there is a growing number of customers going to a check out, so they can automate opening of more shopping lanes and so on. In the event, the user case was to help shoppers in finding all the items and so on, a small screen could easily be added to navigate the shopper.”
The main reason that Raval and the team at Cambridge can develop these technologies with a view to commercialising them is that the sale of the sensory is on. In other words – or in an attempt to recover from that terrible pun – all manner of technologies are falling in price, all because of the ubiquity in the market. As shop managers might like to say, “Pile ’em high, sell ’em cheap”.
Raval says: “Today the CPU power, low-power connectivity, sensors, NFC and so on, are all at a price point that makes them all feasible to be integrated within trolleys and other devices and appliance. I can imagine shopping trolleys that allow you to scan your food as you put it in and automatically charge you on your card when you leave the store with the goods and totally eliminating the need for a checkout.
“By the way, we should not just limit our imagination to trolleys but also think about all the assets that sit in hospitals that have wheels – beds, patient monitors, trolleys and so on. The ability to locate the nearest one has huge benefits from a productivity and a cost perspective.”
The end of everyday human activity
If smart trolleys and other things keep getting smarter, then congregate on the internet of things and launch a revolution, then there’s probably no way to stop them taking over. Which would probably mean fewer humans on the streets, and in their homes. As it is, even in today’s world, a lot of people are too stressed out by technology to learn how to use their contactless payment cards, and prefer to ask the shop assistant to explain what money is. So it’s probably inevitable that shops themselves will go the way of the Amazon and Ocado – and go entirely online.
There are many who believe that it’s inevitable that virtual assistants and mechatronics systems will replace much of what we regard now as “everyday human activity” such as shopping, with the robots even making the payments and managing your money. There are not many people who have not at least shopped for something online in the recent past.
“I am happy to do all my shopping online,” says Raval. “I know that this is not the normal case. Shopping is a social activity and I think people like to see and to experience what is there.”
There are robots now that can manage entire warehouses, not to mention smaller shops. However, stacking shelves in the way many grocery shop workers do is probably still beyond the steadily encroaching mobs of mechanised menaces.
“Yes, I have also heard that stacking shelves is a challenge, especially to ensure that the shelves do not run out,” says Raval. “I have been giving this a lot of thought and we have been working on machine learning and machine vision algorithms to solve these types of problems.
“I am planning a press release on a disruptive machine vision technology for car park occupancy sensor in the near future and would welcome an opportunity to share this with you nearer the time.”
When will end? When will the machines stop? Not until the very last job that we thought only a human can do is done better by one of them. Obviously.