We believe that looking at or processing data feeds one at a time doesn’t help you solve real life problems. Most often you need to combine data from several different sources in order to understand what is really going on, take smart decisions, make really useful, useable and non annoying applications.
Let’s say you have a proverbial temperature sensor in each of the rooms of your house. If you can only treat the data of each device separately:
- You will have one chart per room to look at which is tedious to follow on a everyday basis.
- You will not be able to decide whether you can lower or heighten your boiler’s intensity because you will not have the big picture of what’s the temperature in your house as a whole (unless you calculate it by hand)
- You can’t see correlations between inside and outside factors
- If you want to share your data, you will be sending one notification (for instance a tweet) per feed, which is the surest way to annoy everybody receiving these notifications and turning the information into noise
- You will not turn this info into knowledge, such as, for instance, what’s the equivalent carbon emission cost of your level of temperature.
At Sen.se we consider that being able to combine and process data from different sources at once is a key feature. This is why we have developed several tools to do this:
Lets you get data from several data sources, process them in real time and output just one value. For instance, you can have several temperature sensors in your house. Funnel will get the data from all of them every time a value is updated, then calculate the average temperature in your house. It can also perform other types of predefined calculations.
Select all sorts of numerical data sources, then write your own calculation formula combining the values of all these data sources or fixed values. The simplest use is for instance to convert your energy consumption into Euros or Dollars, but you can also create your own ratios and custom measures.
Select several data sources. Write a text message inserting variables that stand for the data from these sources. Text Mashupper will create a single message using all these values. You can for instance create a single daily report sent by email that summarizes what is going on in your house by showing the values of all your sensors in one message. You can also send a single tweet per day combining your weight (as measured by your bathroom scale), as well as the number of steps and calories burned as measured by your fitbit.
Graph up to three sources of numerical and even text data on the same chart. Multiviz helps you detect the correlation of events and values, highlight how events have effect on one another. Examples are limitless: combine your power consumption and the average temperature in every room of your house; combine the variations of your weight with the number of steps you make daily; look for correlations between the number of people you meet and the number of coffees you drink and your mood level.
As we already said, the purpose of Open.Sen.se is not just to provide a way to store your data somewhere, but to help you make actual use of it. To make use of it, data must be presented in an easily graspable way, not become annoying in the long term, help you learn something, let you make decisions and take actions whether by yourself or by automatically acting on other devices. The purpose, in short, is to solve actual real life problems and not just make technical proofs of concept. If data is not turned into knowledge and action, it is just a waste of energy.
You could already share the data you are posting or generating on Open.Sen.se: make your chosen Sensemeters public and give people the URL of your public Senseboard and voila. The Sen.se/demo public Senseboard is done by using this feature.
But now there’s more. You can display your chosen Sensemeters anywhere you like as long as you have access to the html code of the display page. Post it on your blog, your project website, on onsite public displays, your school or company website…
To do this select “Embed elsewhere” in the Sensemeter’s local menu. You will have access to the customizable and ready to use code to cut and paste in the html code of the page on which you want it to appear.
Here’s an example of a real time Sensemeter displaying all the data we are pouring in our demo account. We pasted the code in our Tumblr based blog post.
You can use this feature to display the values measured by your devices and other data sources of course, like in this graph of hourly average temperatures measured in Montreal.
But you can also use input Sensemeters to give your users posting capabilities or control on your data. For instance, here’s a Manual Logger Sensemeter, through which users can comment about the current temp in Montreal. Their postings appear as dots on the above graph (Put your mouse cursor on a dot to read the content). Try writing a comment.
“Your current power consumption is 534 W”. So what ? Why you need to enrich incoming data on IOT platforms.
Let’s see why you need more than a WYPIAYG (What you post is all you get) platform to make meaningful and useful use of data posted by your devices.
In this post we are going to discuss the first reason: WYPIAYG platforms only handle and offer the raw values measured and posted by your connected devices.
Raw values are like atoms. They are the basis of everything but it is hard to use them for anything practical. You need the platform to turn this raw data into something which you can make use of in a more meaningful way. On Sen.se when you post a value, you can automatically generate one, two, or ten other values that enrich what you have sent. You can use this newly generated data to create visualizations, trigger messages or actions in a more accurate way.
Suppose you have a basic sensor, one measuring your power consumption. Every few minutes (let’s say every five minutes), your device posts your current power usage (for exemple 534 W). What you are usually able to do on a platform that only records raw data is:
- Draw a nice chart showing that currently your consumption is 534W, and showing what this value was five minutes ago, and five minutes before, and five minutes before and so on.
- Send an email or a tweet saying that your current power consumption is 534 W.
- Use the API to retrieve the value 534 W.
You are just showing the same figure by different means, in different places, but you are not making it more meaningful. It belongs to the human user to read this value and try to figure out what 534 W actually means. The current power consumption is 534 W. So what ?
OK, you can also trigger some “smart” action. For instance
- Put at trigger that will send a tweet when the value posted by your device is above or below a given threshold. So you put a trigger at say, 500 W, and you get a tweet every 5 minutes as long as you are above 500 W.
On Sen.se we offer the ability to process data you are posting. So, not only can you use the feed of raw data your device is posting, but you can also use feeds of data that are generated by the platform Apps. Here are some examples of what you can do in the case of our power consumption measuring device:
- What does it mean that your current consumption is 534 W ? Is this high or low compared to what you usually consume ?
- Our Dynamic Gauge App, displays the last value, relatively to all the values you have ever posted. You know how significant the last value is. Besides, Dynamic Gauge creates feeds in which it posts the values when they get to an all times high, or an all times low. You can trigger an action (send an alert, a tweet, turn on or off a device, whatever) when you reach a new maximum or a new minimum value.
- It is hard to always look at numbers and grasp what they mean in simple terms. The Expressive Sign App lets you turn values into status messages. So instead of trying to figure out the difference between 212 W and 324 W and 786 W, you get clear messages saying that your current consumption is LOW, NORMAL, RATHER HIGH.
- Knowing your power consumption during the last five minutes is not very useful practically speaking. You are not going to change your behavior every 5 minutes. The Count Sum Average App, lets you process the incoming data on the fly and create data feeds of more compact, aggregated numbers. For example, the total number of watts consumed during an hour or a day, the average daily or weekly consumption. It is easier to compare your consumption trends on a daily average basis than on a five minutes basis.
- Triggering an action on a per value basis creates dumb or annoying behaviors. You don’t need to post a tweet or send an email or sound an alert every time a single raw value posted by your device is above a certain threshold, because this means that you might end up sending tweets every few minutes which might be annoying and useless. It is much more efficient to trigger actions when a meaningful change occurs, for instance when your power consumption suddenly gets from 500 W to 1000 W. On Sen.se you can monitor the variation between two values, and trigger actions only when the variation gets to a given level. So you will post a tweet or sound an alarm when the power surge happens, but it will not repeat as long as your consumption is about 1000 W.
We believe that the role of an IOT platform is not just to be a pipe and a storage room with a window to look in. For us IOT is about creating meaningful experiences. Providing ways to transform, enrich, extract meaning from sensor and device data must be at the core of a platform on which to build the Internet of Things.
And so we are a so called data brokerage platform for the Internet of Things like a handful of others. What such platforms basically do is:
- Get a value posted by your Device (or another data source)
- Store it
- Visualize it in some way
- Push it somewhere else
- Allow it to be retrieved
Well, doing this is obviously mandatory and useful. But is it enough ?
For us, this looks as if you had a money box. You put your coins in. Then, one day, you open it and all there is are the coins you have put inside. Nothing else was created during the process. It is just a dumb repository of raw data.
When you try to do really useful things out of your data, you realize that acting on the basis of single values is a little bit short. A single value doesn’t show you the big picture, trends, how it compares to other values over time or other data sources.
This is why we are working hard on Apps. Apps process the data you post on Open.Sen.se and create new data, more useful data than the raw bits sent by a sensor. So you not only get the data you brought to the table but you also get much more.
Through a serie of posts on this blog we will be showing how generating data based on your raw data helps you build really cool, useful, long lasting and meaningful uses.
So stay tuned.
Trying to convince an audience that there is probably a life after our screen based interactions era, will be a tough mission, specially at a conference dedicated to screens. But rafi Haladjian, CEO of Sen.se takes the challenge. He will be participating at the Multi Screen conference in Paris on December 13th.
Details on the conference: http://www.benchmark.fr/catalogue/forum/programme/571-forum-multi-screen/
rafi Haladjian, founder of Sen.se and long time pionneer of what is now called the “Internet of Things” will be the opening speaker of the one day conference “Internet des Objets, la révolution silencieuse” in Paris on December 6 2011.
He will be talking about Sen.se’s vision, explaining why we should not focus on the Internet of Things and rather look at the big picture.
Other speakers at the conference include Vlad Trifa from the Web of Things blog, Usman Haque founder of Pachube and Cédric Hutching from Withings.
Details about the conference can be found here.
The 2011 Next conference will be held in Berlin on May 17 and 18. This year’s edition is titled Data Love. What could better suit us. rafi Haladjian, founder of Sen.se will be speaking at the session of wednesday 18 at 15:30 called:Buttons, Behavior, Robots and Toys. What Happens When We Put Data in Things. If you are in Berlin and attending the conference, rafi will be happy to meet you and discuss after his talk.
(This post was submitted by Kamel, we translated his text to english, however his screenshots still bare some french.)
Hello. I use Manual Logger to count the number of coffees I drink in a day. Then I send the result to the Post to Facebook Application. I do the same with the Weather Application. Here are screenshots of the whole process and several views.
Your Device is already posting its data on Pachube ? Do you want to use Feeds of data available on Pachube ? Now we have made it possible to use these data Feeds on Open.Sen.se as if they were connected to our platform. This allows you to use the features, interfaces and Applications of Open.Sen.se with datastreams from Pachube.
Simply go to the Add a Device section and choose Add a Pachube connected Device. You will have to enter your Pachube Key (if you don’t have one, you will have to go and register on Pachube).
To select which Pachube datastream you want to add to your set of Devices on Sen.se you have two options.
- Simply click on List environments and we will list all available public datastreams of all users. You will then be able to browse the list and select the one you want to add.
- Enter search criteria: search for your own or a specific person’s datastreams by entering his username; limit the search to datastreams tagged with specific keywords; look for an environment whose ID you know.
No matter how easy we try to make it to connect a Device to Sen.se (see Automatically generate a Sketch to connect your Arduino for instance) it still takes some time. If you have gone into the hassle of connecting your Device somewhere else on the Internet, we’ll try to get the data from there. We believe that the Internet of Things is not about closed worlds but made of a large number of Device or Data hubs all talking to one another. Otherwise the word Internet in “Internet of Things” would be misleading.
The Quantified Self is a place, a movement, a loose group who gather people who collect data about themselves, their health, their everyday life and try to make sense out of it. QS was founded by Kevin Kelly (former executive editor of Wired) and Gary Wolf (former executive editor of the online version of Wired). QS Show & Tell sessions are held in about twenty cities in the world.
Sen.se is about recording and storing all your data and turning it into action and/or meaning, no matter what the nature of your data is, no matter if it is posted by your devices, some virtual source or manually entered. Self Quantification is one of the things we pretty much believe in and it is one of the purposes of our platform. So it was only natural for us to be a sponsor of the first Quantified Self Conference which will be held at the Computer History Museum in Mountain View CA on May 28 and 29.
For those who are not familiar with Quantified Self, here is a short introduction by Gary Wolf as well as a 5 minutes video of the presentation he gave at TEDx Cannes.