Wireless Wonders

No news, just comment about mobile phones and services, from a veteran practitioner...3G, GPRS, WAP, Bluetooth, WiFi, etc...

Thursday, February 24, 2005

Syncing in the rain...

Do you think that these guys (hypertag) should be speaking to these guys (umbrolly)?


Wireless "data vending" is generating interest. Nokia has a product....

Join my email list
Subscribe to my "100 Mobile Product Ideas" free e-book

Wednesday, February 23, 2005

Umbrella vending...

I just watched a recording of Dragons' Den and saw the pitch for funding an umbrella vending machine. I love it! So here it is....


My mind is spinning with ideas to do with mobiles and wireless and vending!

Join my email list
Subscribe to my "100 Mobile Product Ideas" free e-book

100 Mobile Product Ideas...#16...
...P2P DRM...

We don't want to be like Paris Hilton do we? I mean, talented actress, maybe, but not victims of mobile-data theft (or so we're told). To avoid having our camera-phone photos, address books, or whatever else, being stolen, we need protection.

Moreover, we may just want better privacy. If I take pictures of my family and share them online, then I possibly only want my family and trusted friends seeing them.

The solution to this problem is encryption. In essence, protecting our own data is an issue of digital rights management (DRM). I recently tried to convince a senior Nokia manager that they should seriously consider including OMA DRM on their phones in a way that allows the users to be content producers, not just consumers. In other words, a user could put their own content into DRM-protected envelopes. This is not the intended model, but it makes perfect sense.

Let's say I take a photo of my kids. My device encrypts the photo and then places it into a DRM envelope. That envelope holds a pointer back to my device, identifying it as the place from where to get the key that can unlock the encrypted content.

When a user wants to view the content, their viewing environment requests the key from my device. I can then grant the key, or not, depending on who's asking and any other criteria I may wish to apply.

In effect, users can produce content that is protected in a similar fashion to the super-distribution model. It doesn't matter who copies my content, of if it gets stolen and posted on the Internet. It simply cannot be read until the recipient gets hold of the unlock key, which they can only do by contacting the originator ("content producer") and asking for it. If we imagine a direct contact to the device, thus essentially it would work in a P2P fashion.

Use of DRM in this configuration has some issues. For example, we would need a OMA client on the viewing platform, which might be a PC. As far as I can tell (without much research), the use of DRM for viewing photos on PCs is unusual. However, such a trusted client could be envisaged, perhaps built-in to the underlying OS, or into a browser.

It is interesting to think of using a mobile device as the OMA client too, but working in a configuration that allows the content to be viewed on an associated device, such as a PC. For example, to view a picture (or use other content, such as an address book) on a PC, it has to obtain the rights to do so from the user's mobile device, probably over a local Bluetooth or WiFi link. The mobile then acts as the client for the content and contacts the content producer's device for the unlock key.

This sounds convoluted, but there are some attractive elements to this configuration, such as having an already trusted mobile-to-mobile (P2P) link with the content producer. The need for symbiotic PC-mobile links should not deter us, as this will soon be a commonplace arrangement anyway, for all kinds of reasons and useful applications. In any case, such an arrangement is not necessary, only useful.

Join my email list
Subscribe to my "100 Mobile Product Ideas" free e-book

Paris Hilton...

I'm so incredibly out of touch with the celeb world, that I'm probably at least 10 degrees apart from Paris Hilton, never mind 6. I don't know who she is, but she entered my world when my Google Alerts told me her phone had been hacked.

In itself, hacking Hilton's phone is not that interesting. Being totally non-PC in my assumptions about this talented actress, then I'd say she probably used a password like "hotel". Even if the hacker needed greater ingenuity, someone had already managed to hack the entire Sidekick userbase, making this hack a lot less impressive, at least technically.

However, what it highlights is the lack of proper infrastructure for protecting our data in the mobile world. I think that it's fair to say that address books are highly valued and private collections of information. Users deserve better protection of such assets.

The outstanding hack of the J2ME environment showed how "easy" it was for a rogue MIDlet to gain access to phone book data and call records etc.

In response to that not-so-little Java incongruity, played down by Sun, I posted that handset vendors should be considering encryption of key data as a standard design procedure. This would have defeated the original T-Mobile hacker from gaining access to all those Sidekick-user goodies, including, apparently, their photos and even - apparently - the records of a Secret Service agent (which I find totally incredulous).

Join my email list
Subscribe to my "100 Mobile Product Ideas" free e-book

Tuesday, February 22, 2005

Archival speech recognition...

Following on from idea #14, I did a quick search on speech recognition for archived audio files and found HP Lab's Speechbot. It is currently analysing audio files on the net from various sources with an accuracy as high as 80% (i.e. successful speech recognition).

These kind of accuracies are probably useable, but for analysis of one's own phone calls, the performance would presumably be better (at least for one's own voice). However, with the additional cues offered by meta-data, such as location and who the call was with, overall performance would be enhanced.

Thinking about the call archiving idea, it would be great to be able to extract clips and send them as messages. I'm not sure about the legality of sending clips into the public domain, but I seem to recall that any calls made over a phone network are essentially public (although it is illegal to tap them of course).

Join my email list
Subscribe to my "100 Mobile Product Ideas" free e-book

100 Mobile Product Ideas...#15...
...Stay in touch with the kids...

Hasbro has announced ChatNow, a non-cellular mobile device aimed at kids. My kids already use the Cybiko Xtreme, which is a great device and they think of it as their "Blackberry" like Daddy's. Its user interface is worth exploring. Things designed for kids are perhaps good for adults too!

It occurs to me that if these devices used Bluetooth, or WiFi, then adults could join in the fun. A messaging client on a mobile device would allow messages to flow between the kids and the adults.

Now, some kids might not want the "intrusion" of adults on their airwaves, but I know that my kids would welcome it.

Come to think of it, I'm surprised that a company like Nokia hasn't yet launched a child's phone. Perhaps it doesn't pay to make toys and the real thing?



Join my email list
Subscribe to my "100 Mobile Product Ideas" free e-book

Monday, February 21, 2005

100 Mobile Product Ideas...#14...
...Never forget a thing you said...

The idea of recording phone calls is not new. Wax cylinder recorders were possible in the early days of telephony. What is new, in the age of cheap and dense memory, is the potential to record everything we ever say. Perhaps we can start with our phone calls.

Recording on the device is one option, as demonstrated by Natural Widgets' Natural Recorder. It works on Nokia Series 60 devices and simply records all calls. The files get stored in a circular buffer, so there's no need to worry about file management per se.

Now, the point that I argue in my latest book piece "Mobile in the 3G Era", is that memory is sufficiently cheap that we shouldn't have to worry about file management. Let's face it, who's going to sift through all those files and decide which ones to keep and which ones to trash? It's just like all that stuff that we store up in our garages because we'd rather not decide what to throw.

Everything should be recorded and we simply trash what we don't want after we haven't gone back to if for 1 year, or however long we feel comfortable with, keeping some key files.

I'd much rather see this feature implemented on the network and have all the files spooled to a store on the Internet where I can do what I like with them. I'd probably want to see the files accessible via my mobile personal portal (see idea # 10).

What's essential is an index that means that I stand a chance of finding the file I was looking for. Location and date stamping would be essential and easily done of course. Offline transcribing and novel non-linear search techniques will facilitate searching back through any conversation.

In my book piece I suggest that ALL content produced on the mobile should be perpetually stored. Think of the power of searching this merged content memory. All that data fused with copious meta-data and the power of Web Semantics gives rise to powerful possibilites. "Where was I when I said X to Y?" Who was I with when I said X to Y?" And so on.

Join my email list
Subscribe to my "100 Mobile Product Ideas" free e-book

100 Mobile Product Ideas...#13...
...Give my mum a call...

My Mum keeps a paper address book with all her phone numbers in it. She uses a pen to enter addresses, which, as I posted earlier, is an essential tool missing from most mobile phones. Entering contacts is not easy for people like my Mum, so in effect they are "pen-less". What use is an address book without a pen?

With SyncML, it is perfectly possible to push contact details to many models of phones. Why can't an operator call my Mum and offer to put the numbers in for her, which she simply reads from her address book? Such attention to detail will surely increase loyalty whilst possibly increasing calls on the network.

Join my email list
Subscribe to my "100 Mobile Product Ideas" free e-book

Mobile Avatars...

A quick search reveals that these guys (Veepers) are already doing mobile avatars...look's interesting, though it doesn't have the real-time element.

Join my email list
Subscribe to my "100 Mobile Product Ideas" free e-book

100 Mobile Product Ideas...#12...
...Be who you wanna be...

Avatars are machine-representations of humans and they exist in cyberspace. An interesting commercial use is Sitepal, which I find fascinating. You create an avatar who welcomes and assists visitors to your website. Check out their demo - it's impressive. It might take off, but at a much lower price.

Way back in the early 90's, I was conducting research into the use of avatars in mobile communications. I think that the time and technology has arrived to make this idea a reality.

Imagine making a video call, which perhaps you have already tried on one of the 3G carriers, like Three (UK). Personally, I find that as a substitute for ordinary voice calls, the video element is not that exciting or necessary.

However, I think that the use of avatars can make communications more interesting. The first use is to use an avatar as a pseudo-persona. I make a call and talk as normal, but the recipient sees my avatar and watches it speak the words I utter. I had intended that software could build an avatar that is a caricature of the user, so it looks like me at the other end and it mimes to my voice as I speak.

When I first thought of the idea when I was doing the research, the application I had in mind was online multiplayer gaming. You could see the avatars of the other players. It would be possible to represent one's self as any personality and with any voice from either gender. Originally I had thought that users could chose well known personalities, such as film stars, and use these as their avatar forms, including vocal reproduction.

With 2.5 and 3G networks, the possibility of using an avatar exists because of the support for IP-based realtime communications. Avatars could also be used as messaging envoys to speak voicemail messages or text messages using text-to-speech. The TTS engine could use my vocal signature, so that the speech synthesis still sounds like me and not "Johnny Robot."

With real-time control of the avatar, it would also be possible to alter its form during the conversation. I could change my avatar's hair colour and style as often as i liked during the call. I also envisage being able to convey emotion, giving a whole new meaning to smilies and the use of emoticons. Now I can really look angry, or happy, not just iconically. Perhaps the use of video can control the conveyance of emotions. If I really smile, so does my avatar.

Sooner or later, someone is going to implement this. When I did the research, it was mainly to look at compression techniques to convery the speech and avatar control over a low bandwidth pipe. Today, the compression isn't so important and the technology exists to implement this idea commercially. At the time, my academic supervisor suggested that I change research topic, as there was insufficient prior art to ground the research. Over ten years later, research is no longer necessary. It seems that this idea could be implemented and made usable.

Join my email list
Subscribe to my "100 Mobile Product Ideas" free e-book