Checking and Monitoring Devices Remotely: IoT or historian?

IoT

,

Digital Advisory

Checking and Monitoring Devices Remotely: IoT or historian?

Sara Cavinato | Sep 22, 2020

In this article, we will discuss some of the best technological solutions for monitoring, checking and transferring device information remotely, and try to lay out the principal guidelines for making an informed choice.

In a study from 2011, Cisco estimated that by 2020, the number of connected devices would have reached a record of 50 billion. Growth actually turned out to be slower than predicted, but even if estimates were a little optimistic, we are still talking about a predominant and rising trend.

One other highly topical issue is remote asset management: Edge computing, the concept of incorporating computational capabilities into devices themselves, is a growing phenomenon which first appeared in business and industry, and has now spread to household and even wearable objects.

The evolution of types, complexity and the sheer number of monitoring devices is guiding the development of technologies used to control them.

 On the one hand, there is the tried and tested generation of historian systems, which have been helping manage and monitor activities across various industries since the 1980s by collecting data from the field and making the acquired time series available to users in near real time.  On the other hand, Internet of Things technology has developed in recent years - this technology was invented around 2011 with the intention, from the offset, to become the new generation of control systems, so much so that Gartner highlights it as one of the highest-potential emerging technologies.

The current trend clearly favors the IoT over historians, but is this really the winning trend? 

That depends! In order to discern which solution is best for which of your needs, let’s look at the differences between the two technologies. 

 

Main differences between Historians and the IoT

There are many features that distinguish the new generation of IoT systems from traditional historian technology, from the number of asset features to the complexity of the analyses that can be run. 

Here we give an overview of the major differences between the two technologies, analyzing the most relevant ones as follows. 

 

  HISTORIAN IOT
Scope of applicability Every industry that needs to historicize and analyze time series Industries, smart metering, products (home automation, wearable)
Architecture Monolithic architecture Monolithic or microservice architecture
Scalability Reduced horizontal scalability Increased horizontal scalability
Integration and compatibility Typically ODBC drivers, .net drivers, more complicated integration with UNIX environments. The latest versions of historian have begun to use REST API. Typically REST API. It is not compatible with ODBC and .net drivers
Communication protocols OPC DA / OPC UA / Modbus / DCS vendor registration MQTT, CoAP, AMQP, DDS, http, LoRaWAN (via LoRa Network Server), OPC UA, etc.
Integration with big data (kafka, etc.) Custom Tipically out of the box
Costs and price model Perpetual license costs are high and linked to the number of integrated sources and/or the number of points stored, not the volume of data The cost of IoT platforms is linked to the volume of data and/or devices
Data usage FAT Client, App web, Excel Addin Web app, Mobile app
Analysis Out of the box time series (Trend, XY, correlations) analysis Customized analysis for our of the box time series and ability to connect platforms with business intelligence products (e.g. Power BI)
Number of devices Typically small number of gateways, each reading multiple points Multiple gateways reading multiple points (industrial protocols, etc.), devices with few variables
Response times Typically fast (second) and uniform over time warm storage (past few days, fast), cold (full history, slow)
Claiming No Typically yes
Roles Roles and groups Multitenancy with custom hierarchies
Data structure Time series, typically timestamp value tags, possibly organized according to logical views by equipment Complex Entities and Big Data
Assessments Medium/low complexity rule engine High-complexity out of the box rule engine
Adherence to industry standards audit, cfr 21 part 11 and cGxP compliance audit

 

Data types 

Historians are not compatible with complex data structures or with unstructured data management, but are optimized for simple historical series. In fact, historians have the same analysis performance, typically sub second query speed, on all available history. 

On the other hand, beyond simple time series, the IoT allows unstructured data management as well; think of photos, user and maintenance booklets, and audio, for example. However, IoT performance in historical series analysis is lower than with historians, at least as far as analysis over long-term intervals (months, years) is concerned. 

The IoT typically has two levels of storage with different response times:

  • Warm storage (response time less than one second): this level includes recent data (typically from the preceding month) to allow rapid analysis in the short term.
  • Cold storage (response time of several seconds): Here is where all the historical data that is needed and can be used for machine learning, advanced analytics and more, is kept.

 

Scope of applicability and use cases

As shown in the comparative overview, historians have remained tied to an industrial context, while the IoT has a much broader scope - both of which, however, cover industry. 

Historians are typically suitable for use in: 

  • Plant performance monitoring where the number of assets is limited (e.g. Oil & Gas)
  • Energy monitoring of units or plants for one or more production sites
  • Acquisition of production data for wind farms or photovoltaic plants 

Areas typically suitable for IoT technology are: 

  • All areas with multiple points of measurement of limited types, like the electricity distribution network, for example
  • All industrial sectors in which using wearables is important, e.g. chemical sensors, sensors that measure noise level or the thermal stress operators are subjected to 
  • Sectors where monitoring the distribution network is important, e.g. electricity distribution, distribution of petroleum products at points of sale

 

An informed choice of technology

Ultimately, the choice of technology for managing and monitoring devices remotely will depend primarily on the type of analysis you want to undertake, and the context it is applied in.

As in many other areas, there are no right or wrong choices at first: the most important thing is to have a clear idea of which different technologies are available, and what your needs are, so that you can make the most informed choice. For this reason, we always recommend our clients to begin any digital transformation project with technology onboarding, a preliminary training session that will include from the beginning all the relevant people, and help you to get off to a strong start.

 


Other articles of this series on Digital Advisory:

 

New call-to-action

Subscribe!