Chandra’s Blog

May 14, 2015

I decided to checkout Docker on Raspberry Pi.

There is an image available for Raspberry Pi with Docker.  I started with that.

http://blog.hypriot.com/getting-started-with-docker-and-mac-on-the-raspberry-pi/

The procedure was quite simple.  I downloaded the zip file, unzip-ed it and used a Mac to save the image to a SD Card.

A test program – simple httpd server (mentioned in the above website) also worked great

docker run -d -p 80:80 hypriot/rpi-busybox-httpd

A free way to send SMS to a phone for free using textbelt.com

http://textbelt.com/

https://github.com/typpo/textbelt

— this is how I used it in a demo —-

$.ajax({
type: “POST”,
headers: { ‘Access-Control-Allow-Origin’: ‘*’ },
url: “http://textbelt.com/text”,
data: “&number=” + “4086058889″ + “&message=” + “Occupancy has exceeded the limit”,
success: function(info) {
console.log(‘sms sent’ + JSON.stringify(info));
},
error: function(info) {
// Or log error if POST fails.
console.log(‘Ajax POST failed:’+JSON.stringify(info));
}
});
// test sms facility textbelt.com

I have been using and enjoying Node.js, especially the part about asynchronous programming.  For a hardware trained person, node should be a thrill.

https://nodejs.org/

From their website – “Node.js is a platform built on Chrome’s JavaScript runtime for easily building fast, scalable network applications. Node.js uses an event-driven, non-blocking I/O model that makes it lightweight and efficient, perfect for data-intensive real-time applications that run across distributed devices.”

Also express was my choice of framework to use node and just came across hapi – a newer and more interesting framework.  hapi – Human Application Programming Interface

http://hapijs.com/

From their website – “A rich framework for building applications and services. hapi enables developers to focus on writing reusable application logic instead of spending time building infrastructure”

A nice comparison of express and hapi framework written by  Jonathan Glock  and is available here:

https://www.airpair.com/node.js/posts/nodejs-framework-comparison-express-koa-hapi

I built some simple http server for Internet Of Things applications and so far it looks very simple and easy to use.

In continuing with my experiments with IoT,

I was googling for ways to send mobile phone data (battery level, GPS location) to a centralized server using MQTT.   The nice thing about a MQTT broker is that it is light weight (compared to other message bus and is well tuned for sensor data; further, it also has some QOS as part of the protocol).

Step 1:

I setup a Mosquitto based MQTT broker  (on an Ubuntu 12.04 laptop)

http://mqtt.org/

sudo apt-get install mosquitto
sudo apt-get install mosquitto-clients

su /etc/init.d/mosquitto start
su /etc/init.d/mosquitto stop

Testing:

to subscribe – mosquitto_sub -d -t hello/world

to publish – mosquitto_pub -d -t hello/world -m “hello”

Step 2:

I looked for a mobile app – fortunately OwnTracks (http://owntracks.org/  and https://github.com/owntracks/ios seem to be the best choice).  It is a small app (also available for Android) that sends geo-location to a MQTT server (set up above – all this would be inside a private network).

I downloaded the owntracks app from Apple store and configured it to send data to the server above.

Note:

The MQTT server does not store any information.

On the laptop – mosquitto_sub d -t owntracks/chandra/iphone

On the iphone owntracks app – choose Host ip address (laptop address), default port (1883) and Topic owntracks/chandra/iphone

A quick way to test is to press the square blue button on top (Location Monitoring Mode) and click on ‘Public Now’ button – this will send out a MQTT packet to the MQTT broker (you will see it on the laptop window that is running mosquitto_sub program)

April 13, 2015

I did some experimental setup for playing with sensor data streaming to the cloud – Internet Of Things.

One that seems quite interesting is IBM’s IoT Foundation - A fully managed, cloud-hosted service that is designed to simplify and derive the value from your IoT devices.

https://internetofthings.ibmcloud.com/

IBM offers a 30 day trial and it offers Device registration, Device and application connective, Secure receiving of data, secure sending of commands to devices and Storage and access to historic data.   The 30 day trial is for their IBM Bluemix cloud services.

It is possible to validate their backend with some simple experiments using their Quickstart - https://quickstart.internetofthings.ibmcloud.com/#/

Here is what I tried with Raspberry Pi (B+) - https://developer.ibm.com/iot/recipes/raspberry-pi/

Step 1: Install the software
curl -LO https://github.com/ibm-messaging/iot-raspberrypi/releases/download/1.0.2/iot_1.0-1_armhf.deb

sudo dpkg -i iot_1.0-1_armhf.deb

make sure the service is running – service iot status

to stop the service – sudo service iot stop

to uninstall the package – sudo dpkg -P iot

Step 2:  To visualize the data

Find the MAC address of Raspberry Pi
service iot getdeviceid

This will return something like this:

pi@SensorTag ~ $ service iot getdeviceid
The device ID is b827eba5b0aa
For Real-time visualization of the data, visit http://quickstart.internetofthings.ibmcloud.com/?deviceId=b827eba5b0aa

By going to the quickstart.internetofthings.ibmcloud.com you can see the data send from the raspberry pi.  This works nice.

——–

The data is sent from raspberry pi using MQTT PROTOCOL.

More information is available at the above web site on how to register to IBM’s IoTF organization and register various devices.  It states that once we follow this path, we can send commands to a registered device  (reboot raspberry pi)

Back to sensors…..

Serial to Network – ser2net

http://linux.die.net/man/8/ser2net

sudo apt-get install ser2net

edit /etc/ser2net.conf

Using socket programming, you can access the serial port

Projects:

The NIST Smart Space Project  (Since 1998)
SmartSpace Laboratory - NASA

http://www.nist.gov/smartspace/

Basically they are trying to create a smart meeting room using various sensor
technologies, distributed computing, networking, speech processing, human identification,
and natural language processing.  So far, it seems to be a single room Application.

They have a number of published papers and a user documentation.
(NIST Data Flow System II - User's Guide - Draft Version -
http://www.nist.gov/smartspace/ndfs2_doc_index.html)

They are open to collaborative research and development.  They have HP Labs
along with many others (IBM, BBN, Fuji/Xerox, SRI) as their early collaborators (1998-2000).

P.s. I need to find out (read) more about this project (current status) and their findings.

http://www.usenix.org/publications/login/2000-7/features/smart.html
(Interview article - reference Jini - old stuff)

-----------------------

NRC Smart Spaces Laboratory - Nokia Research Center

http://research.nokia.com/research/labs/nrc_smart_spaces_laboratory

From their web site -- "The increasing smart digital environment requires
of new solutions and services to guarantee the interoperability of devices
and the communication between devices, people and their context. The Smart
Spaces laboratory researches on an array of themes and applications that
include the remote management of our homes, the tracking of our wellness
and the creation of new ways of working together utilizing smart environments."

They are focused mainly on user experience, interoperability, interaction
and collaboration.  Seems to be mobile phone centric; however, they are referring to some interesting application scenarios such as wellness/healthcare.

They have a number of published papers. 

P.s. I plan to review some of their papers (relevant to our project/lab) and summarize.

http://www.w3.org/2007/02/dmdwa-ws/talks/sathish.pdf
---------------------
Smart Spaces Center to research dynamic, independent living
Penn State 

Not much is published material here.  Seems to be a new project focusing on
smart living spaces for aging Americans.  It is a inter disciplinary project
involving psychology, sociology, medicine, business and information sciences.

http://live.psu.edu/story/29755
http://ieeexplore.ieee.org/stamp/stamp.jsp?arnumber=04446560

--------------------------------

BlueSpace - IBM Research

Not exactly a Smart Space -- however, it is   (in their words) "a next-generation
workspace solution ... Integrating sensors, actuators, displays and wireless networks into architectural elements, ...  To increase knowledge workers' productivity by deterring unwanted interruptions, improving awareness and fluid communications".

Their key focus seems to be Concentration, Collaboration and Personalization.

Their long term stated goal is to create a modular workspace solutions that can
be combined in different ways to optimize workspace utilization and worker efficiency.

About 11 people on the project - They seemed to have done a prototype with SteelCase.

They have a number of papers - most recent is dated 2003 (!)

P.s. I need to find out if this is a live project?

----------------------------

Global Smart Spaces
Distributed Computing Group - Trinity College, Dublin

http://www.dsg.cs.tcd.ie/taxonomy/term/65

Seems to be an active research effort. From their web-site -- "Global smart spaces
extend the vision of pervasive computing, in which everyday objects communicate and
collaborate to provide information and services to users, to large geographical areas.
They extend the notion of objects cooperating in a home or an office to the level
of towns, cities, and even countries by integrating a variety of sensor-based and
other systems to provide truly pervasive context-aware services. Such global smart
environments will be heterogeneous as they likely will comprise a multitude of sensors,
networks, and ultimately systems."

They have been periodically publishing their work. Some of them are relevant for our work. 

They also seem to be doing some collaborative work.  For example with Intel, they
have a project called Hermes: A Software Framework for Mobile, Context-Aware Trails.
They have extensive documents and code to download.  I am sure the HP Mediascape folks
will be aware of this project. 

http://www.dsg.cs.tcd.ie/hermes

==============================================================================
Papers:

1. Ubiquitous Smart Spaces:   -- Georgia Tech
http://www.cc.gatech.edu/fce/pubs/darpa.uss.98.pdf

2. Perception, recognition and integration for interactive environments
PRIMA - INRIA 2007 research project activity report
http://ralyx.inria.fr/2007/Raweb/prima/uid0.html

(Seems to be comprehensive, but limited to Perception and recognition.  It has
a few practical scenarios - assisted living, context aware video acquisition,
and augmented meeting)

3. Configuration Challenges for Smart Spaces - USC Technical Report - 1998

Dated and some of the challenges may not be relevant today.

http://research.cens.ucla.edu/people/estrin/resources/conferences/1998july-Heidemann-Estrin-Configuration.pdf

4. Collaboration Applications in Smart Spaces
   MIT Artificial Intelligence Lab

 A high level overview of their project, focused on Intelligent room

http://www.ai.mit.edu/research/abstracts/abstracts2001/intelligent-working-spaces/16oh.pdf

5. Modeling Human-Computer Interaction in Smart
Spaces: Existing and Emerging Techniques -
Telenor Research and Innovation, Norway

http://intechweb.org/downloadpdf.php?id=5535&PHPSESSID=qi0m437u2faqkijjttds4kuqm6

6. inSpace: Co-Designing the Physical and Digital
Environment to Support Workplace Collaboration
- Georgia Tech, IBM, Steelcase

A 2003 paper with some good information, challenges and findings.

http://www.gvu.gatech.edu/reports/08-03.pdf

7. From the Pervasive 2005 conference:

THE UBISENSE SMART SPACE PLATFORM
http://www.pervasive.ifi.lmu.de/adjunct-proceedings/demo/p073-076.pdf

SCALABLE CONTEXT-AWARE SERVICES - Thesis summary
http://www.pervasive.ifi.lmu.de/adjunct-proceedings/doctoral-colloquium/p145-151.pdf

INTERACTING WITH FEDERATED DEVICES
http://www.pervasive.ifi.lmu.de/adjunct-proceedings/doctoral-colloquium/p153-160.pdf

DEMAC: A DISTRIBUTED ENVIRONMENT FOR
MOBILITY AWARE COMPUTING
http://www.pervasive.ifi.lmu.de/adjunct-proceedings/doctoral-colloquium/p115-121.pdf

Dec. 23, 2008

Got the Arudino Ethernet board.  Attached it to Arduino Diecimila board.

Set the ip[]={ 192.168.1.77 };

Used the Arduino-12 software and ran the Sketchbook-Examples-Library-Ethernet-WebServer.

This responds to the web request by displaying the following:

analog input 0 is 301
analog input 1 is 316
analog input 2 is 312
analog input 3 is 299
analog input 4 is 311
analog input 5 is 321

Reloading the page, shows the new values of the ADC values.

Reading up on more stuff on Sensors, detection and context

“Examining Task Engagement in Sensor-Based Statistical Models of Human Interruptibility” – James Fogarty, et. al.

http://www.cs.cmu.edu/afs/cs/Web/People/jfogarty/publications/chi2005.pdf

This is a paper on interrupting at inappropriate times a office worker.  They discuss a way to automatically extract features from low level events (in a restricted work environment)

Related older paper (2003)

“Predicting Human Interruptibility with Sensors: A Wizard of Oz Feasibility Study”

http://www.cs.cmu.edu/afs/cs/Web/People/jfogarty/publications/chi2003.pdf

Sensor based prediction of interruptibility.