Developing Your First Application for IIoT

In this chapter, you will learn about the S95 standard (https://www.isa.org/isa95/) and how it applies to the exchange of asset data. We also provide an overview of databases to store asset data and a comparison of time series storage. Then, we demonstrate how to build a Node.js application that simulates the reading of time series data from sensors and how to enable analytics and visualizations of the data reads. Finally, we provide instructions on configuring Grafana alerts with visualizations, and email and Slack notifications about database values exceeding limits.

This chapter covers the following topics:

• S95 standard and exchange of asset data according to the S95 standard
• Choosing a database management system to store asset data
• Comparison of the most common time series storage
• Using InfluxDB as a time series storage
• Building...

Developing and modeling assets using the S95 standard

The S95 standard, otherwise known as ANSI/ISA-95 (and in Europe IEC 62264), was developed by the International Society of Automation (ISA) to provide guidelines on integrating an enterprise system with a control one. The standard is applicable to all industries and all types of processes within them—batch, continuous, and repetitive.

S95 consists of the following parts, describing different aspects of integrating an enterprise control system:

• ANSI/ISA-95.00.01-2000, Enterprise-Control System Integration, Part 1: Models and Terminology: The document overviews the object models that can help to identify the information to exchange and provides standard terminology
• ANSI/ISA-95.00.02-2001, Enterprise-Control System Integration, Part 2: Object Model Attributes: This part of the standard relates to the attributes of the objects...

Selecting a storage

Each manufacturing facility contains a huge number of assets, from screwdrivers to machines. To track all of them, observing the S95 requirements, you need a suitable database storing asset data. Available storage options can be subdivided into the following groups.

Relational DBMS

A relational DBMS is a table-oriented store based on the relational model (originally described by Edgar F. Codd). Such databases comprise a number of records (entities), each corresponding to a row in a table. The records (entities) contain attribute values.

Relation schemas within a database table are defined by the table name, a fixed number of attributes, and fixed data types. The schemas are the product of normalization...

Time series storage

When we need to store information describing the operation of assets at certain moments over a period of time (for example, information from asset sensors), the best choice is a time series storage. Currently, there are dozens of such databases to choose from. The table here features a comparison of the most widely used ones:

	DalmatinerDB	InfluxDB	Prometheus	Elasticsearch	Graphite (Whisper)
Data types	float62, int56	int64, float64, bool, and string	float64	string, int32, int64, float32, float64, bool, null	float64
Data compression ratio	1	2.2	1.3	22	12
Precision	millisecond	nanosecond	millisecond	millisecond	second
Ingest performance (metrics/sec)	3m	470k	800k	30k	300k
Data source	TCP (binary protocol), OpenTSDB (text), Graphite (text), Prometheus (text), Metrics 2.0 (text), InfluxDB...

Creating instances of assets and adding time series data

In this section, we will build a Node.js application that simulates reading of time series data from sensors. To store the data, we will use the InfluxDB time series database.

For building the application, we will need the following software:

• Node.js 6+ (https://nodejs.org/en/download/)
• restify (https://www.npmjs.com/package/restify)
• node-influx (https://www.npmjs.com/package/influx)
• slack-incoming-webhook (https://www.npmjs.com/package/slack-incoming-webhook)
• Docker (https://docs.docker.com/engine/installation/

To build a Node.js application that simulates reading of time series data from sensors, proceed as follows:

1. Create a package.json file, describing software dependencies. If you use a cloud, specify localhost and mydb as the credentials for the host and database accordingly:

{
  "name": "timeseries...

Understanding the analytics

Big data analytics aims at revealing patterns, correlations, and providing other valuable insights based on processing of large data quantities. By capturing and analyzing big data—large and varied datasets—businesses can better understand their workflows, customer behavior, market trends, growth opportunities, and so on.

Previously, to take an informed decision, a company would have to first gather data, then analyze it, and integrate it into the decision-making process. The whole cycle could take quite a long time to complete. With big data analytics, businesses can speed it up, which gives them the competitive advantage of being more agile.

It is possible to distinguish the following four types of analytics:

• Descriptive analytics: It helps you to get a historical overview of events over a given period and prepare the resultant data...

Deploying your first analytics

In this section, we provide a selection of examples, illustrating how to apply some of the InfluxDB analytical functions to process the time series data from sensors collected using the Node.js application we built earlier.

To enable the analytics for the time series data from sensors read with the Node.js application, proceed as follows:

1. Run a git clone command to copy the source code of the application:

git clone https://github.com/Altoros/iot-book.git

2. Start the application, running the following command:

docker build -t timeseries  
docker run -p 8080:8080 -it --rm --name timeseries-container timeseries

The return will be as shown in this screenshot:

Console output after starting the Node.js app

3. Now, you can start querying data with analytical functions. Data will be retrieved, using InfluxQL, which is a SQL-like query language for interacting...

Visualizing time series data and charts

In this section, we describe how to enable visualization of the time series data from sensors collected using the Node.js application we built earlier.

To prepare for visualizing the data from the application, perform the following actions:

1. Run the following git clone to copy the source code of the application:

git clone https://github.com/Altoros/iot-book.git

2. To start the application, run the following command:

docker build -t timeseries  
docker run -p 8080:8080 -it --rm --name timeseries-container timeseries

Visualizing time series data with Highcharts

Highcharts is a popular library for representing information in the form of charts and gauges. We will use it to visualize the...

Visualizing the outcomes of the analytics as alerts

This section outlines the procedure for setting up the notifications about database values exceeding preset limits. Such notifications can be received in one of the following ways—by email or via a Slack channel. Additionally, you can configure alerts in Grafana.

Configuring email notifications

To configure email notifications, proceed as follows:

• Specify the credentials (email address, host) of the SMTP server
• Change the alertFrom value in the index.js file to a required limit value

Once you have completed this configuration, you will start getting emails, whenever some of the values exceed a specified limit:

An alert displayed in a mail client

...

Summary

In this chapter, we looked into how the S95 standard addresses asset modeling and exchange of asset data between different levels of an enterprise control system. Then, we overviewed the existing database management systems (including blockchain) with a focus on the data types they can store. Furthermore, we compared the most widely used time series storage.

We provided a detailed instruction and a sample code to build a Node.js application that simulates reading of time series measurements. Touching upon different types of analytics, we highlighted the analytical functions of the InfluxDB database. Then, we demonstrated how to enable descriptive analytics with InfluxDB for the Node.js app we had previously built. In addition, we showed how to set up time series visualizations with Highcharts and Grafana and to configure email and Slack notifications and alerts from Grafana...