Portable Context Device: an Open Source Idea

A week or two ago, I drew a picture in my sketchbook of my dream workspace. It had a big draftsman's table with an in-table display as well as a wall-mounted display. Up in the right hand corner of the table, I threw what looks kind of like an iPod with the words "portable context device." Here's the original picture:

Here's a little more on the idea:

I think that a portable context device will be a small, biometrically-protected (uses your fingerprint to prove it's you) device that lets you take your "digital identity" with you between technologies. This means you could sit at your workstation in the office and the device would've already set the box up to match your preferences and your needs, while enabling you the permissions granted by that network to you as a user. It means getting into your car and having the seats and steering wheel automatically get in the right positions, the car audio device automatically configured for your preferred method of operation. It means virtual networking, either with trusted friends and groups, or with people out in public when you're in a public space. I think this is very do-able.

The information the device passes would be context: your name, your telephone number, your IP address at present, all the floaty information that connects temporarily to make up your current proximity (GPS information, etc). There would be all kinds of contact and profile information (mostly provided by you, but I could see you permitting the context device to be some kind of aggregation of lots of data from lots of sources). And a layer or two down in the security would be context about your financial records, your health records, etc. (All protected via encryption layered by biometrics,etc.)

None of this is un-do-able today. The hardest part of the effort would be agreeing on standards, and though I admit that isn't trivial, it's certainly a lot easier than coming up with the technology would've been a few years back.

Look at the "content stack" part of the drawing. I show the different layers of both systems and trust in a "stack" format (which is how people talk about TCP/IP, telecommunications, and other multi-system technologies). So, if you've got this thing in your pocket (it's about the size of a cell phone or a remote control), and you sit down at the local coffee shop, the device will negotiate with the shop's free-floating Wi-Fi service, will register you as present, and will give you the option of being "seen" or "unseen." This is at the public level of context. Those choosing to be unseen will have access to ONLY their own information. Those who opt to be "seen" make themselves available for "pings" by people around them. Thus, if you want to ask someone at another table, "Hey, that looks interesting. What are you reading?" there's a way to do so.

Further down the stack are trusted connections. If you're sitting in the office and a colleague from your work team wants to locate you, they could "ping" you and get your location, open a dialogue with you (voice or text or whatever), share information or files or links, and generally do things with you that you'd allow. Friends and groups you belong to would fall into the Trusted category of operation. Your home network would be trusted.

When you pay a power bill online via your device, this is a secure transaction. Any time you swipe money to purchase something, you'd be entering the secure mode. This would involve heavy encryption, all the bells and whistles of private e-commerce. So, no longer are you trusting the bare internet or your ISP. You're using your device to be the only reliable way to transact that information.

The private region is actually where you store all your own private data. This could be your bill payment records, your dental records, your insurance options, your health care needs. This would be yours and yours alone. Instead of trusting online organizations to keep your data, YOU keep the data and just beam it to their applications through the highly encrypted secure zone to their systems. Private is private.

The File System and Operating System level aren't much to talk about. They control the moving of files, and the operation of software against that device. I think these play the very least in the design, except that we'd require a whole new class of applications to run for these context devices. Skype could become the application of the day, or it might be the app that properly navigates the land-line telecommunications and wireless telecommunications space, and delivers your communication to you wherever you are. (Yes, I realize this sounds a lot like the various one-call solutions, but this is more automatic).

One last thing of note: I like the "radar" idea as I have it. Imagine walking through a mall (do people go to malls any more?) and passing someone in your friends network. They're someone from a group you joined on sustainable living, and you've only met once before. Your radar would blip a little color that signifies a friend is near. You can click this blip to check their credentials, and if you wanted, you could stop and chat. If not, you could at least drop a friendly "ping" to tell them you're nearby. The same would work at the office. When physically close, the radar could ping you that other people in your work team are near. You could set the pings to be more contextual: if you're working in a specific document and other people are working somewhere in the friends network on documents or files that are similar to what you're doing, radar could blip you and tell you this, in case there's an opportunity to collaboriate.

I can't think of any part of this design that isn't really do-able today. I think it lies mostly in the "interface" between all the different systems, and getting the systems to all talk together, but isn't that really what Web 2.0 is all about? Isn't that what XML is supposed to do for us? Or web services?

What are your thoughts on this? I'd love to hear from you.

Technorati Tags: identity , web2.0 , opensource , context , gps , bluetooth , social , xml , annotated