Some of my favorite blogs these days are the ones that track the progress of technology that's been promised to us in science fiction. Blog names like Where's My Jet Pack tell you all you need to know about the site's viewpoint - and it's one I happen to share.
Not that I particularly want a jetpack per se. But I understand the sentiment. I'm not saying I expect lightsabers or a self-assembling racing bike or an honest-to-God Mars colony to become a reality by the end of the decade. But after spending a few days at this year's Computex show in Taipei, I am thinking of starting my own site - Where'sMySmartDust.com.
Mind you, the only smart dust I saw that week was on a PowerPoint slide from NXP Semiconductors. But the fact that I saw it at all was encouraging.
Smart dust, if you're not familiar with the term, is a concept that essentially involves very tiny low-cost wireless sensors operating in densely distributed networks. Put them anywhere, and they automatically detect each other, configure themselves into a network (possibly something simple like a star topology to start with, then later something more complex like a cluster tree) and then sit there and detect whatever they're programmed to detect - motion, heat, humidity, CO2 levels, light intensity, whatever - and forward the data from one node to another until they get to a central server that analyzes the data.
What would you use it for‾ The presenter - René Penning de Vries, NXP's senior VP and CTO - offered a scenario of an intelligent home. It measures the physical conditions of the house or flat, and it knows when you're home and when you're out. It knows how to adjust the A/C and the light switch dimmers seconds before you walk in. It'll also be able to monitor your vital signs (heartbeat, blood pressure, body temperature, etc).
Real-time electric bills
No, de Vries wasn't promoting a new smart dust platform from NXP - he cautions that this sort of technology is still in the university R&D stage, and will be for quite a few years. The smart dust scenario was intended to make a larger point about the importance of innovation.
A more down-to-Earth scenario with wireless sensors - and one we're already starting to see in primitive forms - is the AMR (advanced meter reading) scenario where sensors armed with ZigBee are networked into a home gateway and tell you just how much electricity you're using at the moment - complete with a link back to the utilities company. You can watch your electric bill tick away in real time based on whatever rates are in force at the time (yes - in future, electric companies will offer peak/off-peak usage deals like cellcos do now). The object, says de Vries, is to make users more conscious of how much energy they use so that they can better learn to conserve it.
Even that is a bit future-tech for a few reasons. One: the industry has yet to hash out a common OS for wireless sensors (although the IEEE is working on that). Two: this is hard technology that requires serious network intelligence - especially as network topologies become more complex. Three: no one's really sure that there's actually any money in it.
Still, it was refreshing to see concepts like this being promoted at a show where people seemed more interested in MIDs, particularly the new Asus eeePC line.