REGISTER | LOGIN
Breaking News
News & Analysis

IoT Terrain Still Shifting

Tech landscape painter speaks at ESC
2/27/2015 00:01 AM EST
3 comments
NO RATINGS
Page 1 / 2 Next >
More Related Links
View Comments: Newest First | Oldest First | Threaded View
spike_johan
User Rank
Author
Industries need to drive their requirements (not the vendors)
spike_johan   3/1/2015 1:06:37 PM
NO RATINGS
Rick: Another great article on the IoT - thanks. And this one was especially thought provoking as in my opinion two of the more important issues surfaced here.

First, what caught my eye was when you opened with "Competing platforms are still evolving and pathways to interoperability between them need to be built..." Which got me thinking about how the vendors are all racing to the table with solutions - complicating the landscape - when it should be the individual industries collaborating to drive their requirements back to the vendors. This appears to be another classic case of the tail wagging the dog.

Second, Transport - WiFi, 4G, whatever - is of at best a tertiary issue as the transport solution(s) should be driven from the customer/industry requirement space. Note: Some of the IoT, e.g. sensors and controllers in all/part of the electrical grid should be airgapped. That data doesn't belong in the cloud nor should their control systems be accessable from a smartphone.

Lastly, just like how the US goverment has separated public infrastructure into 16 different classes for security reasons, adds to the argument that the whole IoT space can also be parsed into classes (and sub-classes) bound by the specificities of their requirements.

For example at the very simpliest: we could view the IoT as comprising of three major classes: consumer, business, and industrial. Then for each class (and sub-class) break down the pieces into their constituent moving parts. Then articulate their security, transport, processing, and storage requirements.

A modular, sensible, and logical classification scheme for the IoT would go a long ways towards simplifying the landscape.

Gustavo Litovsky
User Rank
Author
Bluetooth Low Energy
Gustavo Litovsky   2/28/2015 1:09:01 AM
NO RATINGS
With all due respect, the assertions here are wrong. Sure, I may be a bit  since my company works with companies integrating Bluetooth and Zigbee.

 

The IoT terrain is shifting and Zigbee may be falling in the cracks.

The first issue I see is that "The focus of Zigbee, a set of spec that have evolved over the last several years, is on low power nodes that may operate for years on coin-cell batteries — something WiFi cannot enable." Well the thing is that's exactly what BLE can do and what it was designed to do. Now Zigbee is shifting to do it as well? Doesn't bode well for it (although to some degree Zigbee was always billed as lower power, but not Coin cell operated). Once Bluetooth v4.2 starts really rolling out with mesh (perhaps 2016), Zigbee will face competition from a mesh based solution.

Even today we're connectng devices to the internet with a Gateway (much like Zigbee).

Then there is an assertion that "People are focusing a lot of attention on Bluetooth and WiFi, but they are expensive and power hungry," is very incorrect. BLE has been proven to be more energy efficient than Zigbee.

"In "the battle of the network layer," IPv6 is not suited to simple devices like LED lights" except that Nordic just showed IPv6 on their nRF51822. There are companies out there with LED lights running on BLE.

I don't want to hate on Zigbee. It has its strong points, but it seems to me they're trying to make Zigbee into something it's not, and at the same time making BLE seem as if it's not good at what it is actually extremely good (low power).


BLE is very consumer focus, that's true. But it's getting into other things and with Mesh it will grow. Consumers also know BLE and like it. That's a tremendous marketing advantage. Sometimes the technical aspects are not enough.

I did find the analysis of the 8051 switch interesting. As an example, the CC2540 is 8051 based. But a lot of embedded devices are moving to ARM. So, the CC2640 that TI just announced is ARM Cortex M0 and M3 based. It is more expensive (being 32-bit and requiring more circuitry), but it's also more modern and the ecosystem is growing in a way that 8051 just can't meet.  We've seen a lot of companies moving to ARM devices. In a way if you don't do ARM you may not survive except in very niche applications and where cost is extremly critical.

joyhaa
User Rank
Author
zigbee vs others
joyhaa   2/27/2015 11:57:06 AM
worked on zigbee in the past and the experience is not good, IMHO it's hopeless for IoT.

Like Us on Facebook
EE Times on Twitter
EE Times Twitter Feed