![The 2016 Social Video Forecast [Infographic]](https://lh3.googleusercontent.com/blogger_img_proxy/AEn0k_ut44-uuVvSeS0Jnj3NXlnETURekYf7VmFsT2_agBkIbivkF8fJQ6ATzi0sQbT6SkiOFUE8tzXfQw=s0-d)
Imagine you’re walking around Restoration Hardware when you spot it: the dining room table of your dreams. It’s unique, extremely well-made—and just too expensive to buy right that second. So, you find the big red “Pin It” button on top of the table, press it, and move on. Pressing that button just saved the table to your Pinterest board; tonight, you’ll look at it again and decide whether it’s worth the price tag.
[author_more]
Sound a little far-fetched? Well, this hypothetical scenario isn’t actually hypothetical. Pinterest and Tok&Stok, a Brazilian furniture retailer, partnered on an in-store campaign that lets shoppers save physical products to their digital accounts with, well, the push of a pin.
There’s an obvious difficulty. If you push the “Pin It” button on your dream table, and then I walk up three minutes later and push the button again, the button must not only recognize that two different people activated it—but also know which two people, so it can save the table to our respective Pinterest accounts.
The answer comes to us from IoT technology. Each pin uses a Bluetooth Low Energy (BLE) beacon to communicate with nearby consumer smartphones, telling it exactly who’s in the vicinity.
When it comes to the Internet of Things, this use case is just the beginning. Let’s explore how IoT will impact how products are marketed and sold.
Beacons (or, small, wireless devices that send out signals with their unique IDs) are every marketer’s best friend. No, really: not only do they have an unbelievably long list of potential uses, but they give us accurate, comprehensive, nuanced data.
Just like a website heat map tells you what’s grabbing your user’s attention and what’s being ignored, beacon technology can tell you where in your store shoppers are lingering and where they’re rushing. Since you’re getting this information for individuals, you can combine it with other customer data to build incredibly accurate and granular consumer profiles. As an example, you might discover that the average man in his mid–30s spends three minutes in the produce aisle, while the average woman in her mid–30s spends five minutes.
You can collect meta-data as well. For instance, a major clothing store installed beacons in 20 of its locations around the San Francisco Bay Area to track which zip codes consumers were coming from. At the end of four months, the store identified its highest and lowest-density areas.
That brings us to the second use case: data optimization. The clothing retailer used the zip code data to focus its bus advertising on the highest-density neighborhoods; plus, it identified instances where stores were drawing large numbers of customers from a distant zip code, opening the door for future openings.
Beacon technology also lets you send people extremely customized notifications. That’s how Target is doing things: shoppers who opt in to the beacon experience (by downloading an app) get push notifications based on where they are in the store, their personal consumer profile, and product info. Imagine a teenage girl walks into the shoe section. Her previous purchases suggest she likes athletic shoes in the $40 to $60 range, so Target sends her a 10% coupon for a nearby pair of $50 sneakers.
When Hillshire Brands rolled out a similar campaign for its American Craft sausages, purchase intent and brand awareness increased by 20% and 36%, respectively.
That brings us to my favorite beacon use case: giving your customers a good time. That’s how Rockbot, a virtual jukebox that lets customers request songs at restaurants, hotels, and gyms, used beacons. After analyzing historic listening patterns to figure out which songs its users liked best, the platform would automatically play those songs when the consumers walked into its partner venues. Picture this: you open the door to your fitness studio, and your favorite pump-up song starts blasting. Pretty awesome, right?
(And while the ROI of this use case might seem less obvious, there’s a lot of research proving that the physical environment can be as or more important than the product when it comes to driving sales.)
There’s a good chance you already use geofencing in your day-to-day life. Maybe you’ve got an important errand to run the next morning, so you set up a reminder in Evernote, Apple Reminders, or Todoist to go off when you leave the house. Or perhaps you’ve enabled your phone to go on Mute when you arrive at the office.
These digital recipes use geofences, or “virtual fences placed around a physical location”. Every time a person enters or exits the defined area, a programmed action takes place.
You might assume beacons and geofences would conflict with each other, but they’re actually compatible technologies. Since you can define beacon signals down to the centimeter, they’re optimal for incredibly precise interactions. Geofences, on the other hand, can range from 50 meters to the entire city.
Continue reading %11 Seriously Creative Ways to Use IoT for Marketing%
The Internet of Things is all the rage right now. There are so many ideas we can put into action in the realm of physical computing, it is easy to be drawn into the idea of programming the world we live in! Once you have a Raspberry Pi and a breadboard, what's next?
In this article, we will explore how to access the GPIO pins on a Raspberry Pi using Node.js. With the GPIO pins, you can program the hardware directly. JavaScript APIs make this seamless. The APIs are abstractions to common techniques, and they are available from anywhere. The Node.js interpreter runs in a single process which opens up ways to write this code in a way that it is testable. The most exciting part for me is that you can write unit tests, hit breakpoints and examine code just like any other JavaScript program, all from your computer.
Let's get started.
GPIO stands for General Purpose Input / Output. They are the pins found on the side of the Raspberry Pi, next to the yellow video out socket. Below is what they look like.
[caption id="attachment_134480" align="aligncenter" width="1000"]
Think of them as the way you connect to the outside world from the Pi. This enables you to write programs that do not run on a computer screen. Each pin acts like a switch that you turn on or off. You can receive input from the physical world, or send output. Basic boards come with 26 pins, and 9 of those pins are power or ground pins. The ground pins are at the end of each circuit that the current has to flow through. The more recent Raspberry Pi boards come with an extra set of 14 pins.
If you are interested in more details on the GPIO pins, this online diagram gives you all you need to understand what each pin is for. There are a myriad number of pins for input / output and ground. These pins are the foundation of physical computing. Depending on your goal, you can use as many as you need.
fs!I know what you are thinking, what the heck is fs and why do I care? In Unix-like operating systems, a device file is a driver that looks like a file. In laymenʼs terms, a device driver is a file! Guess what? GPIO APIs are wrappers that read or write to a device file. The file system APIs are concepts that may already be familiar to you. If you have never worked with files in Node.js, I recommend going over the fs module and file systems in Node.js. fs is shorthand for "file system" and enables you to read or write to a plain old file. There is nothing fancy here, all we do is writeFile(), for example, and let GPIO handle the rest. The trick is to know what to write on which file.
There is a handy little npm package called mock-fs that will help us with unit tests. With this library, one can dream up any file on the file system and mock it in memory. What is so radical is we are only dealing with files, thatʼs all we need to do. In a Unix-like system, GPIO behaves like any other plain old file. This gives us freedom on how we can approach this solution.
The crux of the mock-fs library is the mock({}) function. It takes in a single parameter which is a JavaScript object. Inside this parameter, one can dream up whatever file you want. The beauty here is that this all runs in memory, so you can go crazy with unit tests. The interpreter runs in a single process, this means one can override the fs module at runtime. JavaScript is a dynamic language, so we are free to mock any module available to the current process.
Continue reading %Getting Started with the Raspberry Pi GPIO Pins in Node.js%
|
The world wide web has continuously developed and morphed throughout its 27-year history. How we display and style content has gone through various phases throughout this time, the most recent of which was the emergence of a responsive web — one web that adapts to fit multiple browsers and device sizes. When it comes to a web with an ever growing number of "Internet of Things" (IoT) devices such as Raspberry Pis, smartwatches, personal assistants that work solely via voice, and so much more — the potential of the responsive web is going to be tested.
The IoT means even more varied devices out there in the wild that could potentially try to load content from the web. We need to ensure that the web remains as accessible and usable to those devices if we are to avoid another big segmentation like we had in the early battles between mobile and desktop. The last thing we want is one web designed for the IoT and one for the mobile and desktop web.
A lot of developers don't quite realise the potentially broad number of devices that are likely to try to display web content to people in the future.
[caption id="attachment_134174" align="aligncenter" width="970"]
Devices such as Raspberry Pis with tiny screens from 2.4" and higher have Wi-Fi, ethernet and even 3G/4G connections, giving them plenty of potential for displaying web content. This content may come in the form of web pages from the world wide web, or it may come in the form of local web pages for embedded device manuals, web-based device control pages and local dashboard style web page displays.
[caption id="attachment_134176" align="aligncenter" width="800"]
Smartwatches can have even smaller displays, with a whole range of different color ranges and screen resolutions. While the Apple Watch doesn't have a web browser yet, Android has a Web Browser for Android Wear by Appfour and there's every possibility that over time more smartwatch web browsers may be on the horizon (assuming someone cracks how to design the app in a way that's easy to use!). At the very least, if the web is to be truly responsive and accessible everywhere, we should be ready for this possibility. What happens if in the future someone is reading an email on their smartwatch and they click a web link? A simple web browser client on a powerful smartwatch isn't too farfetched.
[caption id="attachment_134178" align="aligncenter" width="800"]
One new trend in computing is the smart voice assistant, such as Apple's Siri, Amazon Echo's Alexa, Google Now and Microsoft Cortana. These assistants' role in the home will be to turn lights on and off, set alarms, adjust the thermostat temperature and so on. These services also respond to voice commands and questions with data they retrieve from various places online — usually this involves pairing each company's databases of information with the details it can get about the person using the service itself. For those interested in the area, I've covered how to get started building your own simple artificial intelligence assistants here at SitePoint in the past, including a piece on Five Simple Ways to Build Artificial Intelligence in 2016 and a series on How to Build Your Own AI Assistant Using Api.ai.
Over time, with improvements in artificial intelligence and machine learning, these assistants just might scour the web for information and read it back to us.
To allow this to be possible, the web needs to be accessible to bots who will plan on reading the information out, rather than displaying it visually. Ideally, web pages should already be structured to suit this for screen readers, however adoption of accessible markup is often an after thought today. The emergence of personal assistant AI services just might help change that.
The computer displays of the future, when it comes to the Internet of Things, could include a range of new and pretty fantastic possibilities. Think about devices like smart mirrors, smart windows, smart car dashboards — these will occur just casually throughout our lives, but how nicely will the web display on them? A whole range of new design decisions arise — to display content on a mirror/window, you might need to have more contrast between colors. Mirror/window displays could have a limited color range which web developers will need to account for? It's going to be a whole new (and exciting) world.
[caption id="attachment_134187" align="aligncenter" width="710"]
Continue reading %Front-End Development in an Internet of Things World%
