Friday interview: HP Labs' Huw Robson on how Memory Spots will revolutionise your digital media habits


Tech Digest first wrote about HP’s Memory Spot technology in July, when it was announced. It’s basically a tiny piece of silicon that holds megabytes of data, and can be wirelessly interacted with using a special read-write device. It’s like a turbocharged RFID chip – although if RFID means nothing to you, just go with the ‘computer the size of a grain of rice’ metaphor.

Anyway, it’s got many applications, including medical (embedding Memory Spots into patient’s wrist-bands with full medical and drug records), business (attaching Memory Spots to paper documents with a full record of all the corrections and additions made to the text) and consumer (stick them on your photo prints to add music, commentary or ambient sound when touched with the right device).

I talked to Huw Robson, director of the Media Technologies Lab at HP Labs in Bristol, to find out more about what the Memory Spots could be used for, how long until they’re actually available, and some of the issues around them.

“What we’re trying to do here is wake up storage federated around objects and media in a way that hasn’t been done before,” says Robson. “It was driven by digital photography, and particularly the idea of adding audio to photographs, whether it’s ambient sound recorded at the time you took it, your commentary, music to give you a bit of background, or even a dead relative speaking.”

Frankly that last one’s a bit scary, but you get the general idea. In a sense, it’s doing what you might do with a photo slideshow on your computer, just with a print. When the idea was still mainly photography-related, HP experimented with other methods, including a magnetic strip along the bottom edge of prints, and something Robson describes as “a big fat digital paperclip”, which didn’t work. Hence moving to the idea of tiny bits of silicon.

“The whole question was could you achieve what we wanted to achieve?” he says. “RFID chips typically have a few hundred bits, maybe creeping up to a thousand. It’s enough to give you a unique ID, go off into a central system and use it as a reference to pull back information on that item. So it’s good for stock control, but not for what we wanted to do, storing richer data on there.”

Speed of data transfer was also an issue, as HP wanted to be able to read data off these chips in seconds, not minutes. The current prototype Memory Spots have a 10-megabits-per-second transfer rate, and a storage capacity of up to half a megabyte. And yes, it really is the size of a grain of rice (see left), although for ease of use (i.e. not losing the things down the back of the HP Labs sofa), Robson says the researchers have made them bigger using a suitably high-tech futuristic method.

“We stuck them on a piece of office lamination sheet, and cut them into half-centimetre shapes,” he says. “It shows how robust they are that we can do that. But it also gives a possible use – you could give a user a whole sheet of these things, or have a roll of them inside a printer, which could then be stamped on as you print documents.”

You read or write to a Memory Spot by holding a read-write device (or a phone or PDA with the read-write device in it) millimetres away from the Spot – almost touching it. The Spot receives power through inductive coupling from the device, which can then extract whatever content is stored on the Spot. Having to practically touch it has several benefits, according to Robson.

“There’s no chance of somebody sidling up and reading these things without you knowing,” he says. “And we also have security on board and encrypted keys and all that stuff. Because it’s not reliant on a Wi-Fi network, there’s no lag in getting the data off the Memory Spot. And it’s also positive in another way: touching is so instinctive and natural as far as consumers are concerned. There’s an immediate and powerful simplicity of use.”

One of the issues around Memory Spots is what devices you use to read and write them. There’s two ways of looking at this, says Robson. For commercial and enterprise type applications, HP or partner companies could make dedicated Memory Spot read-write devices, which could be PDA-like devices, or wands, or whatever suits that application.

However, if Memory Spots are ever going to be used for the more consumer-friendly stuff – the photo soundtracks and so on – the technology to read and write the Spots needs to be in mainstream consumer devices, right? Which means mobile phones and PDAs. Robson agrees.

“For consumers, the ultimate application is if it becomes an ubiquitous standard, like Bluetooth, which is in most of your handheld devices. That requires us moving towards standards to make that work, so it’s a longer time-frame. We’re talking to quite a few companies, but it will take quite a few years to have a robust standard with the buy-in from enough companies to make it take off.”

In other words, while HP could put Memory Spot technology into, say, an iPaq PDA fairly soon, it’s going to be a while before you see a Memory Spot-enabled RAZR phone hit the market. Robson says a commercial or enterprise deployment of Memory Spots could come within a year or so, but it’s likely to be two to three years before the technology finds its way into consumer devices.

Another factor will be cost. It’s hard to say how much each Memory Spot will ‘cost’, partly because it’ll depend on whether tens of thousands of them are being made for an enterprise application, or billions for widespread consumer use. In the former case, Robson reckons a single Memory Spot might be around 50p, but in the latter, the price could come down to pennies.

“Volume means the price can come right down,” he says. “We’re not using any cunning special silicon technology. We’re making them using the standard silicon process, on a wafer, and we can take full advantage of Moore’s Law.”

In truth, some of the most quoted applications for Memory Spots seem a bit… impractical. How many people really want to have a soundtrack for their printed photos? Especially when so many of us now share our photos using email or sites like Flickr. But in many ways, the best thing about Memory Spots is the technology’s flexibility. Robson says new uses are being suggested all the time, some of which seem to me to have more appeal.

For example, what if the Memory Spot attached to your print didn’t have a soundtrack, but instead held the digital JPG of that photo, enabling you – or whoever you sent it to – to get a perfect copy made of the print at some point in the future? Or if every CD in HMV came with a Memory Spot on the front that you could touch your mobile phone or iPod to in order to hear some samples?

“It’s a fascinating bit of technology that’s taken us a bit by storm,” says Robson. “Loads of people have ideas about how we could use it, so we have to whittle those down and pick the most appropriate. We don’t think there’ll be Memory Spots in everything. Well, not in our visible future, anyway…”

Stuart Dredge

One thought on “Friday interview: HP Labs' Huw Robson on how Memory Spots will revolutionise your digital media habits

  • This story about HP’s Memory SPOT has got to be one of the most laughable (or contemptible, depending on your perspective) news stories in awhile.

    There is a guy from the USA’s Massachusetts Institute of Technology, named Irving Tsai, who has to be one of the most, if not THE MOST, ripped off yet prolific inventors of this time period.

    Tsai invented all of the things “announced” by HP in this news story – way back in 1990-1993! Take a look at his patents in the USA and Canada and EPO by the title “Still Film Sound Photography”, “Tape Recording Method and Apparatus”, and his sequence of patents dealing with what he has referred to as “Multimedia Paper”, which was hardcopy that retains all the electronic functionality of the electronic hypertext document it is printed from.

    Tsai spoke of Java-like stick notes, photos with sound, photographic slides that automatically run the projector they are in, faxes that automatically link to the web or show video or dynamic content, and stuff like that more than a decade ago when Netscape was just starting out.

    Tsai’s US No. 7,016,084 patent in fact controls the operation of all modern web browsers, in a way similar to what US company Eolas currently alleges to have control over – though they were later by over a year than Tsai and showed far less. It is clearly a genune tragedy of nature that the most creative minds tend to occur in people who are such social misfits to the point of keeping silent while literally everyone robs them. (And in this case one has to also wonder what role prejuidice is playing, to the benefit of those stealing this inventor’s creations.)

    It is also widely know by people in certain circles that everyone from Bill Gates to Steve Jobs to Michael Dell to even Opera Software’s Jon Von Tetzchner and their counterparts at SUN Microsystems and other companies were “privately” studying the Tsai inventions for things they could use without acknowledging the source.

    So, Patent Offices on both sides of the Atlantic should take a good look at what Tsai disclosed before erroneously granting any patents to HP or others in this area of technology. And any technology historians out there, if you are sincerely interested in truth and historical accuracy, this person single-handedly invented much of what the “giants of industry” are trumpeting, as in this fine article, as “their” marvelous inventions.

Comments are closed.