by Joesph Boakai, Chief Technology Officer at Intellegant
‘Internet of Things’ is one of the new century’s buzzwords, with the potential of creating a new industrial revolution, the so-called Industry 4.0. This phrase was first mentioned by Kevin Ashton of Procter & Gamble [1], then later MIT's Auto-ID Center, in 1999 and in November 2001 [2].
What is IoT?
But what really is the Internet of Things? There are many definitions, some of which are quite confusing. A basic source, Wikipedia, defines Internet of Things as [3]:
“the network of physical devices, vehicles, home appliances and other items embedded with electronics, software, sensors, actuators, and connectivity which enables these objects to connect and exchange data”
Of course, the concept covers far more nuances and applications than that definition implies.
Each “thing” is uniquely identifiable through its embedded computing system and is able to interoperate within the existing Internet infrastructure.
These "things" can be anything from industrial machines, home appliances, vehicles, mobile devices, human body sensors, wearable devices, and other items embedded with electronics and software.
IoT is a key driver for the Digital Transformation – providing the ability to measure parameters never before used, and thus the ability to do new data techniques such as Predictive Analytics.
IoT End-2-End Workflow
Getting deeper into how IoT works, the end-2-end workflow of an IoT implementation is typically as follows:
The Device Interface Gathers the data via smart sensors.
IoT traffic generally communicates using an MQTT/REST protocol – the Send/Submit step. This is also the data ingestion pipeline for the Broker and Storage layer.
A Streaming and Batch analytics layer is used for Analysing the data to gain meaningful insights.
In the Acting step, other services, like Reporting and 3rd Party Interfaces, send alerts/notifications.
The entire end-2-end workflow is Secured by (hopefully) stringent security policies.
This is further explained in the below illustration from Deloitte report on IoT architecture:
Flavours of IoT Technology
IoT comes in many flavours as follows:
Industrial Internet of Things (IIoT)
Networking house lights and thermostats is one thing, but IIoT and the massive potential it has to streamline manufacturing is one of the most exciting aspects of IoT. IIoT involves the integration of complex physical equipment with high-end software and networked sensors for factory automation and predictive analytics. Since there are many different devices used in manufacturing and industry, you need gateways and data converter units to convert the protocols from one to another. These data are collected from various industrial devices, information is processed, and then upload onto a repository/enterprise asset management (EAM).
Of course, some companies many need to first ensure that all their equipment is standardised by retrofitting legacy equipment. For a fully automated implementation, you may need sensor fusion – the combining of data coming from group of sensors, which is more efficient than one sensor at a time. The combined data is then temporarily stored and uploaded to a server, a process called data concertation.
Benefits of IoT Technology and Solutions
According to a Gartner 2017 report [4], “24 [billion] USD will be spent on IoT Endpoint / Sensing Devices. And by 2022 IoT-enabled service model[s] could save a Trillion USD a year in Maintenance and Service cost[s]”. Therefore, the business benefits for harnessing IoT technology are quite clear!
Among many uses of IIoT, a particularly strong one is Predictive Maintenance, a method designed to diagnose and help determine the condition of in-service equipment to predict when maintenance should be performed. This is a clear example of using IoT and advanced analytics to enable better decision making, and, ultimately, significant cost savings. Some other benefits of deploying IoT in industrial settings are:
To gather and analyse operational data using actuators and IoT sensors, visualize the results, and help maintenance engineers proactively identify and manage equipment reliability risks that could adversely affect plant or business operations.
To help recommend optimal maintenance schedules with the use of prescriptive analytics, with the goal of better utilization of maintenance resources.
To help identify operational factors that positively or negatively affect equipment performance, by employing machine learning techniques.
To improve equipment reliability and reduce downtime by making assets and equipment more intelligent.
Using real-time data and analysing historical operational data collected from IoT devices or sensors to predict failure, and allow fixes before damage occurs.
o This approach promises cost savings over routine or time-based preventive maintenance, because tasks are performed only when warranted.
IoT Architecture
Since IoT technologies come in different forms, the architecture deployable for businesses is also different. For example, the IoT architecture used for home automation, home monitoring, etc., is different to the ones used for Industrial IoT applications such as Manufacturing, Smart Grids, or Smart Health.
(IoT Reference Architecture - Source: WSO2 Inc)
At the customer or business ends, the IoT buzzword is used or misused without knowing exactly how to benefit from deploying these technologies within an enterprise IT ecosystem. Not all IoT technologies are suitable for every business mentioned above. The type of IoT architecture and sensors deployed for monitoring and managing Smart Grids or Biomedical applications, for example, may not be suitable for home monitoring applications (and vice versa).
Many customers have been confused by other parties in the best use of IoT technologies for their business, and many are struggling to figure out which type of architecture or enabling IoT technologies to deploy. Finally, they also need to consider how to best leverage the data generated from IoT to gain insights, make better decisions, and add value.
Cloud IoT Platforms
There are many Cloud IoT Platform vendors on the market today, some of which are the big technology giants such as Amazon/AWS, IBM/Watson, Google, SAP, GE, Bosch, and more. These vendors provide both Public and Private Cloud IoT platforms, depending on the need and the sensitivity of the data in question to be hosted.
For the end customer, it is a bit difficult to know exactly which of these IoT Platforms to use and deploy for their business, even before getting into Data Security and Data Privacy considerations. The main differentiators amongst these Cloud IoT platforms vendors are as follows:
Ease of Connectivity – Devices / Sensors / Actuators to be connected to stream data.
Available Tools for development – Advanced Analytics and Real-Time Insight for decision making (including Machine Learning and Artificial Intelligence).
Edge Analytics – Capability to do Edge Analytics for local decision making while reducing data feeds.
Out of the Box Capabilities – Applications and ready-to-use solutions.
Security – Guaranteed security to protect sensitive data and to gain trust.
Cost and Pricing – A Pay-As-You-Go model, prices for data volume streamed, etc.
An example of an IoT Platform is below:
( IoT Platform - Source: IoT World)
Some IoT Challenges
The choices above are complicated by some of the below concerns, some typical of any emerging/maturing technology:
Market Fragmentation: the IoT market is fragmented with many players in hardware, software and services.
Standardization: fundamental issues in low power & lossy networks e.g. IPv4 -> 6LoWPAN/IPv6.
Security Concerns: integration of IoT security with traditional enterprise IT ecosystems.
Data Science Skills: there is a skill shortage in this field, and finding a consultant with all the necessary IoT and Data Science skill can be difficult.
Interoperability: IoT requires that devices seamlessly and directly communicate with each other and the Internet (e.g. M2M communication), and benefits can suffer if there are problems.
Conclusion
Deployment of IoT technology and solutions will bring exciting new business opportunities and benefits. According to Gartner [4] on IoT adoption, soon 20 billion IoT sensor units or devices will be installed in variety of business settings, such as manufacturing, home automation, mobility, healthcare, energy, agriculture, entertainment, and banking & finance. Specifically, IoT is creating new capabilities such as Predictive Maintenance that will help prevent downtime and save billions, and potentially doing it all remotely, with distance monitoring.
In order for businesses to decide on the right IoT technology, Architecture and Cloud Platform, the customer first needs to select the right implementation partner who have the expertise and understands the IoT technology in various industries to do the following:
1. Assess your Business and IT Infrastructure
Assess the As-Is situation and then propose the right IoT architecture to be deployed for the business.
2. Selecting the right Cloud IoT Platform
As mentioned earlier, there are many IoT platform vendors out there and selecting one is a matter of choice. But depending on the type of business and IT Infrastructure, your partner will help you select the one fit for purpose.
3. Start with a LEAN Approach
Don’t start with a big-bang.
Start with few devices, machines or sites then rollout across the enterprise.
4. Support and Maintenance
Your implementation partner will help you with the best approach for support and maintenance after deployment.
Please do not hesitate to contact us if Intellegant can assist with your IoT questions and/or implementation and/or all other smart data solutions issues you should be interested in.
***
Sources:
[1] Procter & Gamble
[2] MIT Auto-ID Center
[3] Wikipedia
[4] Gartner Inc.
Comments