XUELK-AN-005: Implementing an Industrial IoT gateway/supervisor

Info Box Applies to SBC Lynx

History

Version	Date	Notes
0.5.0	May 2017	First draft
0.6.0	May 2017	Added TBD

Introduction

SBC Lynx is extremely flexible in terms of communication interfaces and I/O ports. When used in combination with a Linux distribution such as Debian, it is an ideal solution to implement compact highly-integrated cost-effective gateways/supervisors for Industrial Internet of Things applications (IIoT).

IIoT is a vast field that includes several disciplines. At field level, the typical IIoT system requires

• collecting data from different in nature networks and fieldbuses (an RS485 bus, an Ethernet LAN, an IO-Link port, etc.)
• uploading the data to the cloud over an Internet connection.

This application note provides some examples of hardware and software configurations that can be used for these purposes. Also, it shows how to harness the versatility of SBC Lynx to integrate on the same platform even processing and controlling functionalities. [1]

The examples were tested on an SBC Lynx board equipped with optional wireless module and that runs Debian distribution, as described here.

[1] It is worth remembering that SBC Lynx supports IEC 61131-3 compliant soft-PLC engines as well, such as Axel LogicLab.

Cloud interfacing

At the time of this writing, tens of cloud platforms are available on the market addressing IoT and IIoT applications. [1] Some of them provide free accounts for small projects or evaluation purposes.

The following sections describe how to interface an SBC Lynx-based gateway to some of such platforms.

[1] See for example this section.

Ubidots

Ubidots is a codeless IoT Platform designed to help you prototype and scale your IoT projects to production.

This example consists of three steps. The first shows:

• How to connect the SBC Lynx to Ubidots
• How to upload data to the cloud.

The second part describes how to visualize on the dashboard of Ubidots. Tha last one shows how to set up events so as to get a notification.

The procedure here described was tested on a SBC Lynx running Debian distribution. For more details, please refer to this application note.

Connecting and uploading data to Ubidots

As stated before, it is assumed that the board runs a Debian distribution:

root@arm:~# lsb_release -a
No LSB modules are available.
Distributor ID: Debian
Description: Debian GNU/Linux 8.7 (jessie)
Release: 8.7
Codename: Jessie

To connect to the Ubidots platform, a client written in Python will be used. The client will make use of the Ubidots' client library (https://pypi.python.org/pypi/ubidots/). To install it, please issue the following commands as root user:

apt-get update
apt-get upgrade
apt-get install python-setuptools
easy_install pip
pip install ubidots

On the cloud side, we need to have a user account on the ubidots.com platform and get the API (a unique key for each account ) and token from the website, very simple.

Go to the website of the Ubidots cloud and click the sign up option on the right upper corner of the website to create the account.

To get the API token for the just created account, go to the right upper corner of the home page and click on the username. Then you will see four options which are My Profile, API Credentials, Docs and Log Out.

Click the API Credentials. The API token (fwSetosAgIG3d128YdzP5fi4hADzWU in this example) is a string which will be used later to initialize the API client instance.

To create and initialize an API Client instance, use the following code:

from ubidots import ApiClient
api = ApiClient(token='fwSetosAgIG3d128YdzP5fi4hADzWU')

On the cloud side, it is necessary to create the variables that will store the values uploaded by the SBC Lynx. In the example, three variables are created: Fall_sensor, Flood_sensor and Temp_sensor. To create variables, there are three options in the upper center of the home page of the Ubidots interface, as shown below.

Click the device option, then we will get the following:

Click the device and add a variable:

After adding the variables, we need their IDs. To get them, put the cursor on bottom side of each variable box. Then, you will see "I" option. Click it to get the ID.

To create an instance of the variable on the SBC Lynx side, use the following format code:

my_variable = api.get_variable('<variable ID>')

To upload a value to the Ubidots platform, we can use a code like this:

new_value = my_variable.save_value({'value': <value>})

Following is a code snippet that uploads random data to the platform:

from ubidots import ApiClient
import random
api = ApiClient(token='fwSetosAgIG3d128YdzP5fi4hADzWU')
my_variable_1 = api.get_variable('58cfe99b7625427aeab8a5e8')
test_value_1 = random.uniform(0,1)
test_value_1 = my_variable_1.save_value({'value':test_value_1})
my_variable_2 = api.get_variable('58d0df2f7625427ae8065765')
test_value_2 = random.randint(1,10)
test_value_2 = my_variable_2.save_value({'value':test_value_2})
my_variable_3 = api.get_variable('58d0f2d77625427ae6d4847f')
test_value_3 = random.randint(1,10)
test_value_3 = my_variable_3.save_value({'value':test_value_3})

Visualizing data on the cloud dashboards

To visualize the data sent to Ubidots, we have to use the Dashboards section, where we can configure how we want to visualize the data.

First, click the Dashboards link. Click the "+" symbol on the right upper corner. Then you will get the "How would you like to see your data?" question and different options below it:

• Chart
• Metric
• Map
• Indicator
• Controller
• ...

In this example, the option chart was selected. As such, several further options are available:

For all of the variables, the Line chart option was chosen. The following image shows how the data visualization looks like:

Events and notifications

Ubidots allows sending notifications—for instance in the form of text messages (SMS) or emails—to the authorized users, based on the events configured on the cloud. In the following example, a variable that was not mentioned in the previous sections is used (heartbeat).

To configure notification options, click enter the Events section.

Then sekect the device (My_device_1 in the example) ...

... and click the variable you wish to configure the notification for.

After selecting the variable, we will define the condition that triggers the notification. In this case, it is triggered when the heartbeat variable is less than 4:

Last, we compose the message we want to be emailed every time the notification is triggered.

Device-to-device communications

Configuring SBC Lynx as a MQTT broker

This example shows how to install a MQTT broker on SBC Lynx, running Debian distribution. "MQTT, a simple, lightweight, publish/subscribe messaging protocol on top of the TCP/IP protocol, is the ideal protocol for the emerging IoT world" [1] and is becoming one on the most common protocols used for machine-to-machine communications.

In contrast to HTTP with its request/response paradigm, MQTT is based on a publish/subscribe model, as depicted in the following image. This makes it suitable for "Internet of Things" messaging where even highly constrained devices are used (low power sensors, mobile devices, embedded computers, microcontrollers, etc.).

Publish/Subscribe way of communication is an event-driven and enables messages to be pushed to clients. The central communication point is the MQTT broker (server). It is in charge of routing all messages between the senders and the rightful receivers. Each client (sensors) that publishes (transmits) a message to the broker includes a topic into the message. The topic is the routing information for the broker. Each client that wants to receive messages subscribes to a certain topic and the broker delivers all messages with the matching topic to the client. For more details, please refer to the publicly available resource such as this link.

[1] https://www.ibm.com/developerworks/cloud/library/cl-mqtt-bluemix-iot-node-red-app/

Installing the broker

Eclipse Mosquitto™ broker will be installed on SBC Lynx. Eclipse Mosquitto™ is an open source (EPL/EDL licensed) message broker that implements the MQTT protocol versions 3.1 and 3.1.1.

The procedure was tested on an SBC Lynx running Debian Jessie distribution. First, import the repository package signaling key:

wget  http://repo.mosquitto.org/debian/mosquitto-repo.gpg.key
sudo apt-key add mosquito-repo.gpg.key 

Make the reposioty available for the mqtt:

cd /etc/apt/sources.lit.d/
sudo wget http://repo.mosquitto.org/debian/mosquitto-jessie.list

Install mosquito

sudo apt-get update
sudo apt-get install mosquitto

7. Install Mosquitto developer libraries to develop MQTT clients sudo apt-get install libmosquitto-dev 8. Execute the given below command to install Mosquitto client packages $sudo apt-get install mosquitto-clients 9. Ensure that Mosquitto broker is running $sudo service mosquitto status

10. simple test of the broker Open a terminal and issue the given below command to subscribe the topic “mqtt” 1 $mosquitto_sub -h localhost -t "mqtt" -v Open another terminal and issue the given below command to publish message to the topic “mqtt” 1 $mosquitto_pub -h localhost -t "mqtt" -m "Hello MQTT" Now the message “Hello MQTT” will be displayed in the first terminal where the topic “mqtt” is subscribed.

Now, we can also test the functionality of the installed broker (Mosquito), we will need two client one as a publisher and one as subscriber.

References

List of IoT cloud platforms (not exhaustive)

• mnubo
• Oracle
• Swarm
• Axeda
• OpenRemote
• Etherios
• ioBridge
• SAP Internet of Things Solutions
• Zatar
• ThingWorx
• Arrayent
• Sine-Wave Technologies
• Ayla Networks
• Echelon
• EVRYTHNG
• Exosite
• Xively
• Marvell
• Carriots
• Arkessa
• GroveStreams
• CeNSE by HP
• ARM
• Nimbits
• Open Sen.se
• Paraimpu 15
• Sociot.al
• NewAer
• SensorCloud
• ThingSpeak
• Yaler
• Jasper
• XobXob
• Linkafy
• Revolv
• Wind River
• Wovyn
• InfoBright
• Contiki
• 2lemetry
• AllJoyn
• InterDigital
• Superflux Internet of Things Academy (IoTA)
• HarvestGeek
• MediaTek Labs
• Streamlite LTE
• Bosch Software Innovations Suite
• Geras
• Ubidots
• Opensensors.IO
• Tinamous
• Temboo
• Eyehub
• Nearbus
• Exosite
• Emoncms
• Gemalto
• IBM Watson
• Microsoft Azure
• Amazon AWS
• Senseiot
• Parse
• SICS
• Element blue
• Sparkfun
• Kaa
• Salesforce
• PRDIX