DASH7 Wireless Sensor Networking Blog

This is intriguing.  I had no idea ZipCar had 4.4 million users.  But this new idea of private car owners renting out their own cars to strangers is intriguing once automakers start to include more sensor logging capabilities into their vehicles to detect the kind of abuse that I believe rental cars get on a regular basis.  I want to know if the guy I rented my car to a) drove too fast, b) did a brake torque, c) did donuts on a baseball diamond at school, d) drove on really bumpy dirt roads, e) spilled anything on the seats f) sat on the hood with his friends drinking Budweiser.

I would also be curious what how the car companies are viewing ride sharing services and whether they are much of a substitute for actual car purchases or more a substitute for mass transit or car rentals.

I spent 90 minutes cruising the exhibits at Embedded Systems Conference (ESC) this week in San Jose, CA.  A few nuggets:

- Proprietary wireless technologies/implementations continue to prosper.  802.15.4, zigbee, wirelesshart, wifi, and bluetooth so far have not shaken the industry up sufficiently to dislodge its preference for proprietary solutions.  Unlike other wireless shows I attend, standards were not at the top of the agenda for this crowd, based on my quick snapshot of the event.

- energy harvesting is becoming very real and DASH7’s hyper-low power draw is a terrific match for future “batteryless” apps.

- One 802.15.4 device integrator who gave a class on mesh networking says he begins his class by telling his students that most people who think they need mesh networking, don’t need it.   And the only people who really need it are the ones that are using RF technologies with poor range.  Otherwise, mesh is unnecessary.

- Another 802.15.4 device integrator talked about all the different flavors of 802.15.4 that his company is building products to support and that while interoperability is a theoretical possibility, interoperability is still not a big priority as they see it and of course, 802.15.4 is not so much a single standard but rather a hodgepodge of multiple standards or specs that are loosely related to one another.

Important writeup on DASH7 and interoperability from Ann Grackin @ ChainLink research.

lots of sensors …

but there’s no standard.  We’re starting with TPMS and envision DASH7 taking hold elsewhere in the automobile.  That’s the good news.

The bad news is that someone in Detroit thinks your windshield “display” should one day look like this:

in those books

about how to run good meetings, there’s usually a directive that there needs to be a “timer” for each meeting.  This sort of starts to address this in a humorous way (how much is this #$)(@*)$ meeting costing the company?) but someone could combine this with an access control product and install it in every meeting room in a building.  With a DASH7-enabled badge (note– this would be for enterprise apps only … not consumer!) that looks identical to the 13.56 MHz prox card you already use, your attendance (or absence) from a meeting can be automatically logged, the time you leave can be logged, and there could even be an average hourly cost of the meeting based on the salaries of each person in the room (this is probably a bad idea but I just had to say it).  Chronic meeting lateniks can be logged and outed and can stop making you wait for a meeting to start.  (Natch, if the latenik is the boss, this feature will probably not be welcome either … oh, this idea is starting to grow hair!)

For consulting companies, law firms, and others who bill clients hourly, this could assist with billing.

Would be cool if someone like ADT or Johnson Controls came up with a solution here … seems to me like an extension of existing access control solutions but for a worthy cause.  Layering in conference callers is an added level of complexity for later …

About half of the meetings I attend are not well managed – no agenda, opaque objectives, no shows, no one taking notes, no timer.  If a solution like this could begin to address the lack of a designated timer in each meeting, even through the use of a humorous angle like “how much $ is this )#$()(@* meeting costing us?”, it might be a minor victory in our daily battle with endless meetings.  Could even make some of them more effective.

Oh, and you won’t have bluetooth or WiFi enabled badges since the batteries will die quicker than a zebra doing a solo trek across the Serengetti.

pretty specific words from NXP

Getting NFC onto a critical mass of smartphones is a huge leap forward for the RFID/WSN community.  For the first time, we’re looking at a “ubiquitous” wireless sensor data acquisition platform that can enable hundreds or thousands of applications that aren’t practical today with the high costs of fixed RFID infrastructure or the bulkiness of “mobile” RFID infrastructure used in so many supply chain applications today.  Similar to the way the barcode made the jump to hyperspace decades ago (there are now trillions of barcodes out there, for any of the doubters out there), for the RFID community this is more important than last decade’s Wal-Mart mandate (supply chain) and potentially more important than the various DoD RFID mandates.  Unfortunately for many of those invested in passive “Gen 2″ RFID, NFC represents a different frequency and a (mostly) entirely different set of use cases than they are accustomed to talking about.  For some in the passive RFID community, NFC appearing on smartphones is the equivalent of the opening number at a AC/DC concert (blinding lights … explosions … dry ice fog … stacks of Crown amps blaring @ 120 decibels) overtaking a talented but underpowered John Mayer solo pianist warm up act.   Not trying to diss passive or John Mayer fans here, it’s just that the sheer firepower of the smartphone app developer community, the sheer numbers of smartphones, their global reach, and the marketing firepower of the brands behind the smartphone industry  (think Samsung, DoCoMo, Apple, AT&T, Vodaphone, Nokia, et al)  are just overwhelming to the current crowd of RFID companies that I see at most of the RFID conferences I go to.

The coming incumbency of NFC combined with the very high power draw of Gen 2 passive RFID all but guarantees Gen2’s absence in any major smartphone platform as a reader.  Adding a Gen2 tag to a smartphone is certainly doable, but the notion of Gen2 as a stepping stone to a broader sensor data acquisition platform on smartphones is effectively moot now.  But even the notion of adding a Gen2 tag to a smartphone seems unlikely to get much traction now … two separate, non-interoperable passive RFID technologies on a single device?  For what apps?  The NFC opportunity (mobile payments, ticketing) is obvious and large.  The comparable Gen2 opportunity is not.

The bigger opportunity in RFID for smartphones is around so-called “active” RFID, or as it’s referred to in other circles, wireless sensor networking.  The ability to acquire sensor data from a heterogeneous array of things, places, or people in a diverse  RF environments while at the same time requiring only a trickle of battery power is the holy grail for sensor networks and smartphones.

While WiFi and Bluetooth have become incumbents on many smartphones, their high power draw and high latency (the time it takes to connect with another node) make them impractical for adding to all those heterogeneous things/places/people.  The history of wireless sensor networking is littered with the corpses of technologies that required frequent battery swapping or recharging.  While Bluetooth LE solves some of the battery problem, it’s short range and high latency make it viable for a limited number of apps like personal body networks (e.g. Nike shoe monitoring apps).  Both WiFi and Bluetooth and fine for high bandwidth streaming apps, but for sensor nets, they are just overkill.  Why hook up the water cannon when the squirt gun will do just fine?

There are a bunch of emerging specifications based on the IEEE 802.15.4 standard that have mostly not been published but in any event are not interoperable and do not address the issues of power or latency.  The mess that is 802.15.4 is a topic of a future post but basically the thing to remember here is that 802.15.4 is not really a single standard but rather a orphanage of specifications operating at different frequencies, using different PHY’s and different MAC’s and generally attempting to create interoperability hell on earth.   Zigbee is “based” on 802.15.4, but zigbee does not = 802.15.4.     It’s probably easiest to remember that any standard that starts with the three numbers “802″ is one that was designed as a session-based protocol, not as an asynchronous protocol for sending short, bursty sensor messages.  The people at zigbee have taken 802.15.4 about as far as it can go in terms of promising a solution that addresses the power issue but in addition to being high latency and not being all that low power in the first place (zigbee uses about 10x the power of DASH7), most of their implementations operate at 2.45 GHz.

Which leads me to another one of the trap doors of 802.15.4.  A development unrelated to sensor networking in the smartphone community is the emergence of 802.11n (there’s that 802 again — another session based protocol) wifi and for protocols operating at 2.45GHz like zigbee, 802.11n is like the wireless neutron bomb.  It just destroys other RF operating at 2.45GHz mercilessly and with no quarter.   For this and other reasons, you are not likely to see much adoption of zigbee on smartphones.  I should add that some in the zigbee community are finally figuring this out and scrambling to move some implementations to 900 MHz, which is fine in the USA but outside the USA, not so good.  Spectrum issues.

So putting NFC aside for a moment, DASH7 makes a terrific case for being the “right” wireless sensor networking platform for both smartphones but also the things/people/places that those smartphones interact with.   Long range, multi-year battery life, signal propagation through walls/water/concrete, single frequency, single PHY, single MAC.  There’s more.  For these reasons alone DASH7 is hyper-compelling.  But with NFC, DASH7 becomes even more attractive since DASH7 can occupy the same silicon as NFC with only a tiny increase in the bill of materials.  For the handset vendor who doesn’t want another radio or doesn’t want to further burden his BOM, DASH7 is a terrific story, as it is for the carrier as well.  The first combo DASH7-NFC chips should hit the market sometime in Q3/Q4 this year … I’ll keep you posted on our progress.

this is similar to the History Channel app

I blogged about last week … quite a lot of opportunity here to associate digital content with specific “things”, starting with places in this case via Foursquare.  No reason this can’t be extended to all sorts of content based not just on places but on things (Edmund’s review of the new Infinity M45 Sport), people (protesters picketing outside a store could be linked to content about the beef they are having with the owners of the store), animals (DASH7-enabled dog tag could be linked to an article about a famous kennel where this breed of dog was born).

All paid for by pay-per-click advertising.  Phones can be subsidized, or using the equivalent of AdWords, you could choose your own advertising for each of the examples above.  And share in the revenue with Google.  More exclusive “things” have higher PPC ad fees.  For example, an Absolut vodka ad “painted” on my colleague JP would be priced lower than, say, the same ad “painted” on the cocktail waitress at Strait’s Cafe.  But I digress …

the shape of things to come

Google’s obvious target:  pay per click ads everywhere in your Augmented Reality browser/computing environment/smartphone …

in case you missed this …

As Cattle Rustling Increases, So Does the Need for RFID.