How to Choose The Best Industrial Internet of Things Platform?

In Blog
By

If you’re an experienced technician, businessman or project manager, often it can be frustrating how many options are there on the market. Which platform is the perfect fit for industrial use-case? How much should I research to gather enough data for decision making? We wrote this blog to give you insights of market-leading industrial IoT platforms fast and easy.


The Industrial IoT (IIoT) is the key driver behind a strategic Industry 4.0 initiative, which is going to be massively implemented worldwide by 2020. The fundamental principle of the Industry 4.0 is transforming hardwired machines into smart cyber-physical systems capable of exchanging real-time data, interoperating with each other, and flexibly reprogramming themselves on the fly.

Key metrics for the Industrial Internet of things

There are numerous benefits of the Industrial IoT, but the most important fact is that it drives direct business value based on production data management at the massive scale and near real-time speed. This value comes in four primary forms:

  • Increased revenue through improved asset utilization and the introduction of new, smarter products
  • Reduced operational cost through more efficient energy and resource utilization as well as improved worker safety.
  • Improved customer satisfaction through better performance, reliability, and usability of the product.
  • Improved worker safety through quick autonomous or remotely controlled equipment to decrease safety-related incidents.

The safest and surest formula to start with the Industrial IoT is to take up predictive maintenance first. The implementation cost is moderate, but the impact on revenue growth and optimized production can be huge.

When production processes are interrupted, the company loses money. Even if it is only minutes per month per machine, it adds up significantly if we talk about a large fleet of industrial machinery. For some systems, like a power plant, a military system, or an oil refinery a failure is not an option at all.

In this blog post I will cover what most manufacturers are looking for when they’re choosing an industrial IoT platform, and mostly they are those seven solutions (in order of importance):

  • The Flexibility Of The Platform - Customizable microservice architecture can simplify customization and ensure sufficient separation of concerns between different parts of the platform.

  • Predictive Maintenance - The aim of (PdM) is to predict when endpoint may break and prevents it from happening by performing the maintenance.

  • Predictive Quality - Predictive Quality is about understanding the patterns in data to determine the areas of highest risk and directing resources before the risk becomes a reality.

  • Condition Monitoring - It’s a fundamental component of Predictive Maintenance. It monitors the vibration, temperature or any condition, which may interrupt the high-quality production.

  • Automated Root Cause Analysis - Showing the cause of a causal chain that leads to a specific outcome or defect.

  • Energy Consumption Optimization - Visualized data of electricity consumption from the device to the platforms web-dashboard.

  • Digital Twin - A digital visualization of the physical objects or the system real-time and throughout its development changes.

That’s being said, let’s dive into the platforms.


KaaIoT Technologies

Pros:

  • Flexibility: Customizable microservice architecture
  • - The Kaa microservices interact via open APIs and can be integrated with third-party systems and rearranged according to your needs. You can even skip those microservices that you don’t need or replace any of them with your own components saving on deployment costs. Some other platforms on the market can only emulate real separation of microservices as a result of trying to convert existing monolithic architecture to the modern approaches.

    - Because of customizable microservice architecture, expanding the existing functionality is much easier for industrial IoT and not only.

    - One more advantage of Kaa architecture is that you can add support for any IoT protocol in a matter of days by adding a “plug-in communication microservice”

    - https://www.kaaproject.org/platform

  • Clear pricing
  • - Kaa provides transparent pricing. Kaa-hosted package starts at $750/month at the time of writing. Follow the link down below if you’re interested in self-hosted or embedded pricing structure.

    - https://www.kaaproject.org/pricing

  • On-premises deployment
  • - TCO of the Kaa-based, on-premises solution is much lower than a similar project by a competitor.

  • The platform is not bound to specific protocol, this gives you ability to use any existing protocol with the appropriate plug-in.

Cons:

  • The new generation of the Kaa IoT platform - Kaa Enterprise isn’t an open-source
  • -0.10.0. Avocado Archipelago is an only open-source platform available for now.

  • A wide variety of different microservices and features increases the time and effort to learn the platform

Microsoft Azure

Pros:

Cons:

  • Interestingly enough, Microsoft has decided to explain about their predictive maintenance through only one use-case. I’m not sure how everybody's needs and wants can fit into only one use-case but that’s what they have right now. I was not sure whether to put into the cons or pros but I think having only one use-case and no availability to dig into the technology is big no for me
  • - https://azure.microsoft.com/en-us/features/iot-accelerators/predictive-maintenance/

  • Requires Platform Expertise - Every platform is unique. Some of them are easy to use and adapt, however, Azure requires expertise to ensure all moving parts work together efficiently. Meaning that you will have tons of meetings and calls until everything runs properly.
  • Azure often pushes businesses to put all eggs into one basket. They mostly propose you to accept the single vendor strategy. Often a single vendor increases comfort; however, it comes with a higher danger as well. When one provider manages data and also becomes your launching platform, then what's going to happen if Microsoft is incapable of fulfilling their contracted obligations? Sue them? For some business, it's unnecessary risk to take.

Amazon Web Services (AWS)

Pros:

  • Integration with other AWS services like Lambda, Kinesis, Sagemaker, …
  • Predictive Maintenance/Predictive Quality:
    AWS has a rich set of analytics systems like Amazon Sagemaker that allows running complex analytics on collecting data - https://aws.amazon.com/blogs/iot/using-aws-iot-for-predictive-maintenance/
  • Energy Consumption Optimization:
    Amazon Kinesis provides an ability to visualize IoT data.

Cons:

  • Complex pricing, it’s hard to predict in advance how much will you pay
  • The highest prices across all IoT platforms on the market
  • No on-premises deployment.
  • Hard to build a good visualization (no integration with 3rd party tools like Grafana, poor built-in data visualization tool)
  • Flexibility: AWS limits you to one of the following protocols: MQTT, WS, HTTP.

Cumulocity

Pros:

Cons:

  • SDK-oriented connectivity model forces you to choose pre-integrated devices only.
  • Integration with other AWS services like Lambda, Kinesis, Sagemaker, …
  • Predictive Maintenance/Predictive Quality:
    AWS has a rich set of analytics systems like Amazon Sagemaker that allows running complex analytics on collecting data - https://aws.amazon.com/blogs/iot/using-aws-iot-for-predictive-maintenance/
  • Energy Consumption Optimization:
    Amazon Kinesis provides a good ability to visualize consumption data.

International Business Machines (IBM)

Pros:

Cons:

  • Only IBM Bluemix provides built-in microservices architecture. Bluemix, unfortunately, doesn’t support IoT protocols.
  • -Deploying a custom set of rules can be time-consuming due to the lack of flexibility in IBM Watson (an IoT platform).

  • Unfortunately, they don’t provide automated root cause analysis. IBM asks to gather the information manually and send the general information when reporting a TPC-R problem.

  • - http://www-01.ibm.com/support/docview.wss?uid=swg21292738


Challenges of adopting the IoT technologies

Many businesses are facing difficulties with industrial IoT adoption. They don’t know when and where to start. How may it bring the highest effectiveness and what technologies should they use?

Those questions are spread out, and it is okay to ask. According to Forbes Amazon has one of the best customer satisfaction ratings on the market.

However, most of the reviews come from consumers, so it is hard to evaluate AWS objectively. On the other hand, Kaa takes customer satisfaction rating very seriously, and each inquiry is essential for them. If you don't find exactly, what you're looking for over the internet or other companies delay with the answer, I highly recommend contacting Kaa.

So what are those main challenges, you may ask?


Security: IoT endpoints interact with each other automatically. Executives and technicians don't want to think about predictive maintenance of devices because of the IoT platform visualization. Not exactly, detecting hack attacks or data leaks has nothing to do with the predictive maintenance or data visualization. Keep in mind standalone security elements will have to be introduced in the network at the very start of the adoption.

An expense of deploying IoT solutions: This is one of the most crucial factors why companies sometimes stay indecisive. The cost of implementing the IoT infrastructure often can be devastating. Various organizations worry about ROI and effectiveness of the solution. However, choosing the right IoT vendor can turn things around and run the adoption process smoothly.

Want to know more about using the IoT for your business?

Share on FacebookShare on LinkedInShare on Google+Tweet about this on TwitterEmail this to someone