Smart Technology Series: Episode 2: IoT Protocol Inflation

August 6, 2018 by · Comments Off on Smart Technology Series: Episode 2: IoT Protocol Inflation
Filed under: Business, Finances 

As with every new technology trend, there are always standards that compete with each other trying to solve the new problems encountered by the technology needs.  Back in 2010 when the mobile world was nascent, we had Mobile OS Inflation.  Clever technologies, as we pointed out in Episode 1, are no exception.  In fact one could argue that IoT has broken the record of protocol standards.  As Nordic recently posted on their Get Connected Blog: “from ANT to Zigbee”.

The challenge with “things” is that they have very different needs that will drive different connectivity standards.  Wireless standards are typically designed to optimize battery life while maximizing performance for the specific application they are designed for.  They key performance attributes that IoT devices are typically concerned about are:  range, data rates, frequency of data transfers, latency, security, and reliability.  The “goldilocks” of these requirements are clearly different for different applications.

Let’s look at a few examples:

  • A self-driving car will need long range, high and frequent data rates with low latency but it benefits from large power source. The network needs to be secure and very reliable.
  • A remote thermometer will need long range but low rates and infrequent data transmissions. High latency is acceptable, but battery availability may be limited.
  • A beacon will need to be really small, thus driving a very limited battery capacity, but its data needs are also limited. Depending on the application, the range needs can also be short.
  • A switch needs very low latency, but very infrequent and low data rates, while power is not an issue since they benefit from being hardwired.

These examples make the point that one wireless standard will not be sufficient for optimal operation of all of them.  So the need for various protocols is real.

Ant+, Bluetooth, Zigbee, and Z-wave are very similar in nature but do differ in terms of performance attributes.  Ant+ (owned by Garmin) has been primarily focused on fitness devices like heart rate monitors, cycle computers, etc.  Bluetooth is perhaps the most ubiquitous.  It is present in every smartphone, tablet, and computer originally as a low range low data rate communications protocol. However, recently with the additions of BLE and Bluetooth 3, it is expanding into many more IoT use cases than it was before.  Zigbee is arguably the IIoT (industrial IoT) protocol of choice.  It has been available before IoT was even a thing (no pun intended).  It is a very mature protocol.  Extensions like Zigbee Pro and Zigbee Remote Control give it flexibility with reliability, security, and scalability for very complex systems.  Z-Wave, relatively new to the party, is a very low latency and low data rate protocol that supports meshing, making it ideal for home automation projects that include lights, valves, sensors, and the like, where nearly instantaneous action is expected.

WiFi, of course is everywhere and growing.  Newer variants can support rates up to 1 Gbps.  Chipsets are very power efficient, scalable, and extremely cost effective these days.  It is mostly a communications network protocol, but due to its omnipresence, there are a lot of IoT devices out there based solely on WiFi.

For long range needs there is also a growing collection of protocols for IoT applications like  Thread, Sigfox, Neul, or LoRaWAN.  There is a version of Cellular IoT that 5G technologies promise to make widely available.  It has high data rates and capacity with extremely low latency.

The big question is what technology to use for IoT projects. The truth is that it is not possible to settle on a single one even for a simple smart home project.  We have launched applications that combine several of these in order to have the best possible outcome.

It is unlikely that a single protocol will win the IoT battle.  We can expect extensions to some of the most common ones to make them more suitable for different use cases.  Some protocols may wind up relegated to niches and some may even vanish in the future.  However, we will have to live with protocol inflation given the diverse nature of IoT needs.  Successful installations will focus on the right choice of devices and protocols optimized for each part of the project rather than a choice of standards to make it easier for the installer.  Remember, it is about the best performance of the application, not about technology choices.

Enjoy.

Facebooktwittergoogle_plusredditpinterestlinkedinmail

Smart Technology Series: Episode 1: Is it Smart or Clever?

July 24, 2018 by · Comments Off on Smart Technology Series: Episode 1: Is it Smart or Clever?
Filed under: Business, Technology 

Prefixing everyday things with the adjective “smart” seems like a good marketing practice.  But how truly smart are smart things?  There is no common baseline to call a thing smart.  Not that I am suggesting to regulate the term, the way organic foods are.  Creating a committee of smart people to regulate what smart is, does not seem smart at all.  But it does leave things too open for the average consumer and technologist to figure out.

A few years back, I tried (unsuccessfully, I might add) to coin the phrase “clever phone”.  Are you old enough to remember the Palm Pre and the Microsoft Kin? Calling a device “clever” rather than “smart” makes sense:  Not a simple device, but one that can do a few more things than a traditional device but not quite as smart as an iPhone or Android.   I think we are at a similar point now with the so called Smart things. Being clever means being skilled at doing or achieving something; being somewhat talented.  Whereas smart implies some intelligence, or the ability to do something independently for a positive outcome. In a sense, smart involves some decision making process whether it is natural or artificial.   It is in this bloggers opinion that most so called Smart Things now are just clever. 

Let’s start with Smart Lights:

If you get a “smart light”, like the Phillips Hue, you can do all sorts of fancy things with it.  You can change colors, dim the light, have it turn on with a specific trigger, etc …  but there is one fundamental flaw: the light switch has to be on.  If you are tired at night and just flicked the light off (as most of us do), that is it, all your smart programing goes out the window until you remember to flick the switch back on.  And if you just want your light to go on as you flick your switch, let’s say to look for something, you may not be able to, since the switch just provided power to the light so it can then apply the “smarts” you spent so much time programming.

How do you solve that, you might ask?  Well, the first attempt is to add the smarts to the switch so it controls a “dumb” light.  The problem with that approach is that your light will only turn on, off and maybe dim; no colors here…

So, IMHO, both the light and the switch are clever since they can do more than my traditional switch or light.  But neither is really smart per se.  The way I have worked around this is obviously by combining both:  a “clever” switch controlling a “clever” light.  That way, you can control when the light goes on and off as well as the color temperature and intensity of the light.  You can still flick it on and off as you do with a traditional light.

We can extend this idea to a bunch of other Smart Home things like speakers, appliances, locks, blinds, garage door openers, etc.  They are all limited, but do perform things that the traditional counterparts don’t.  However, the combination of things and good programming can make things smarter. But, installing, setting up, programing, and securing is not for the faint of heart.  The ecosystem of these “clever things” is a mess, a real mess.  There are hubs, protocols, devices, wireless standards, IFTTT, etc.  Just look at the list of supported devices in Google Home.  But that is a topic for a future post.

In essence, calling IoT “things” smart seems like a promise, rather than a reality.  In order to use them to create a true Smart Home, Smart Enterprise, Smart Healthspace, or Smart Buildings you will not only need to mix and match them, but you will need to program the smarts to make them unique.  So let’s just be candid, call the things clever and write spend the necessary time making these installations truly smart and unique.

Enjoy

Facebooktwittergoogle_plusredditpinterestlinkedinmail

Connected World

May 30, 2016 by · Comments Off on Connected World
Filed under: Business, Technology 

cloudofthingsv2As the term Internet of Things (IoT) becomes as colloquial as “smartphone” or “texting”, one can’t help but to think about the implications to modern life.  It is undeniable that it is the biggest economic opportunity in the world since …well, the Internet (although I always considered computers as being “things”).   The concept that devices (aka things) can sense stuff in their environment,  send data autonomously, and act on commands at a relatively low cost is what makes it so disruptive and intriguing.  There is no surprise that companies like Apple, Microsoft, Alphabet (the company formerly known as Google), and Amazon have been gobbling up startups and investing oodles of cash into it.  This is a technological transformation similar to the digital era that changed the way we read, listen to music, or watch movies. Besides these are the companies best suited to drive the trend. There will be hundreds of other smaller companies in this industries developing devices, software, sensors, and network elements.  But these 4 have to be at the epicenter of the action.

In the humble opinion of this obscure blogger, there are 4 elements worth analyzing that are affecting this new trend:  The evolution of technology, the “things” themselves, the data they send or receive, and the system to make the whole thing do something useful.

  1. Technology Evolution: This answers the question of why now.  Moore’s law, Metcalfe’s law, and the development of fast wireless is making it all economically and physically possible. In essence, a lot of processing can be done in very small sizes and at a very low cost.  It is now cost and size effective to have electronics in almost every “thing” to sense the environment and transmit data.  None of this will even be thinkable without the right scale of technology.  The right scale is happening now.
  2. Data:   “Things” sense data and autonomously transmit it.  A “thing” can send identification data, status data, environmental data, location data, actionable data, or a combination of these.  A message saying “the third fridge in the ice cream section is malfunctioning” can be an example. A message like this can be sent to a store operator’s device so s/he can go and physically check the fridge and fix the problem.  But if there is a feedback loop, a system can remotely trigger an action to correct it automatically.  In other words, data can be just informational, actionable, or a command itself. The data is captured by sensors that are meaningful to the object: Temperature for a fridge and weight for a truck are examples.  An intelligent system will then make sense of the data by analyzing it and then propose or trigger an action. Sense, connect, analyze, and do something about it.  Simple, right?
  3. The “things”. These are the “what”.  I like to classify them according to seven major environments where “things” are.  The individual, home, vehicles, public places, the workplace,  the enterprise, and infrastructures.  The data is different across all these and, more importantly, it is used for different reasons.  The first four imply a direct relationship with individuals and they are inter-related.  In all of them, for example, we’d like the temperature to be comfortable so an automatic thermostat may be enough.  An individual may have control over the temperature in his/her car, but not in the mall or the office. So there is no need for these thermostats to interact with each other.  However, if you google (or is it alphabet now?)  the restaurant for your business dinner, wouldn’t you want it to seamlessly show up in your car’s navi so you can drive there, and then on your phone for the last few steps?  Or would you like your playlist to continue from your home, to your morning run, and then your car?  These require an integration beyond just their own environment and the end user needs to be the center of the experience.  The connected enterprise, workplace, and infrastructures are a bit different.  In these cases the “things” are the core of the operations.  My simple ice cream fridge example above is one of those.  These will generate a significant ROI for companies, so they will most likely develop faster.  More on this in future posts …
  4. Architecture.  This is where the debate gets not only nerdy, but interesting.  The industry is toying around with 3 basic architectures: Smartphone centric, cloud centric, or environment centric (based on a local hub or a mesh of objects in the environment).  In my opinion, there always has to be something in the environment managing and connecting all the “things” even when the smartphone is not there, so lets consider that a given.  But there are pros and cons to a smartphone vs. a cloud architecture.  And as it often occurs when new technologies are being launched, the control of the experience is the battleground.  It is in the best interest of the smartphone players to have it be the center.  Apple or Alphabet (through the Android ecosystem) will thrive in this environment.  For the strong cloud providers, like Microsoft or Amazon, cloud is the way to go. But from the user’s point of view a combination is the best option.  So the word “centric” needs to not be that relevant.  Again, companies like Apple and Alphabet have a great presence in both the smartphone and the cloud.  In addition to that, their hubs, AppleTV and Google’s OneHub will complete the offering.  Microsoft and Amazon do not have enough presence in the personal device but are very well positioned with cloud services and hubs, so they will also be strong players.

To paraphrase Dr. Carl Sagan, there a billions upon billions of “things” out there that need to sense, get connected, monitored, and acted upon. Some estimates put it in the trillions of dollars of economic activity. Consumers will benefit significantly with a seamless and open architecture.  What made the Internet so successful is, in part, the openness.  One would hope that the “things” will also be as open.  If this market is to be as big as predicted, we will need a lot of large companies and lots of smaller ones to make it work.  Openness will make it happen.

Mr Cook, Mr Nadella, Mr Bezos, and Mr. Page, I hope you agree when you read this. If you don’t, let’s discuss.

Enjoy.

Facebooktwittergoogle_plusredditpinterestlinkedinmail