When COVID-19 threatened to derail the manufacturing industry, it was Industry 4.0 that saved many manufacturers. Notwithstanding the supply chain issues that continue to pose challenges, 94% of respondents in a recent McKinsey survey said that Industry 4.0 helped them keep their operations running during the pandemic. The connective tissue that makes Industry 4.0 is the Internet of Things (IoT), which connects millions of physical devices to the internet.

However, despite the familiarity most have with IoT, many manufacturers – who admit that the pandemic was the push needed to embark on digital transformation – are still grappling with how to maximize the benefits to digitize their shop floors. 

What capabilities should you look for in an IoT platform? Well, data management sits at the core of every successful IoT project. Enormous amounts of information are exchanged between the connected devices in an IoT system. A good IoT platform does the following:  

• Collects data: Think of all the devices and machines sitting in the shop – data can be gleaned from all of them, and that data can be valuable to many of your other systems.  A good IoT platform is constantly collecting data from a wide variety of data sources. 
• Processes data: Data is valuable, but sometimes data alone is not enough. Raw data typically needs to be run through some sort of processing algorithm to transform the raw data into something useful. A good IoT platform can analyze, filter, and process data to produce useful results.  
• Utilizes data: Collecting and processing data is only part of the job. The processed data must be fed back into the business to create operational improvements for the entire effort to be worthwhile. A good IoT platform disseminates the processed data throughout the shop ecosystem so that the data can be utilized to improve business operations.

What is the most effective IoT system architecture? Well, there are many ways to design and configure an IoT platform, and these options can be confusing. The design dilemma ultimately comes down to where you want to process and store your data.

Cloud Computing Offers Scalability and Cost Benefits

Remember how the cloud was viewed just a decade ago?  Cloud computing was initially met with significant skepticism, particularly from a reliability and a security perspective. Companies simply didn’t trust it. It could go down. It wasn’t secure. Companies were hesitant to move critical and confidential data off-premises and into the cloud due to these concerns.  

To their credit, cloud providers addressed many of the initial issues and continued to evolve the technology, making it more reliable and more secure.  After further investigation and experimentation, companies began to trust the cloud to address the shift to mobile and digital that customers were demanding. And, before you knew it, cloud computing became generally accepted and widely used.

However, cloud technology is still far from perfect. Concerns about reliability and security remain. Cloud outages are certainly far rarer than during the infancy of the technology, but one in five companies reports that they have had a serious or severe outage within the last three years. Such outages are expensive: 60% of outages cost more than $100K, up from 39% in 2019. And companies continue to be concerned about the security risks associated with moving their data outside of their own infrastructure and onto networks and hardware/software systems that are configured, managed, and secured by others.

The primary advantages of cloud computing are scalability and cost.  Despite any perceived weaknesses, the cloud is quite effective at storing and processing big data because processing resources are not limited to any physical on-premises hardware and since large quantities of data can be stored and easily accessed via data lakes in the cloud.

Edge Computing Extends the Benefits of On-Premises Computing

With cloud computing emerging as a hot technology, we haven’t been hearing as much about on-premises computing solutions. However, on-premises computing remains the trusted norm for most companies that deploy and manage their own compute at their own facilities.

With edge computing, you are deploying your own systems as close to the data sources as possible. You maintain complete control over your hardware, your software, your configuration, your interfaces, and your data. Having your applications running on the on-premises edge computing devices allows you to collect, process, and utilize data as close to the data source and as quickly as possible. This low latency allows users to expect more efficient data processing and superior business continuity when edge computing solutions are deployed.

In some ways edge computing can be the exact opposite of the cloud. By utilizing your own solutions deployed at your own facility, you are eliminating internet and third-party dependencies as well as most connectivity and security concerns. Unlike the cloud, where downtime can be a major risk, edge computing is renowned for its reliability. It is an incredibly stable technology that rarely suffers from outages and that typically provides the highest possible overall compute availability (especially when redundant, fault-tolerant computing solutions are utilized). 

Hybrid Blends the Best of Both Worlds

The debate used to be about the cloud versus on-premises computing. The initial thinking was that nothing would ever replace on-premises systems but, over time, companies had to utilize the cloud for cost savings and for scalability. Once the cloud took off, the conversation shifted to whether the emergence of the cloud would kill the on-premises systems. It didn’t. Businesses learned that each compute strategy provided unique advantages that the other did not. Thus, the perfect design was often using the two together.

The same is happening now with edge computing and the cloud. Companies are addressing their most critical data requirements by deploying edge computing solutions to process data at the data source.  But they are also leveraging the cloud for their big data storage and analysis needs. This hybrid approach maximizes the benefits of each computing strategy and results in an overall system that is reliable, responsive, scalable, and cost-effective.

Cloud Computing vs. Edge Computing

When deciding between these approaches, the following factors must be taken into consideration: 

• Security: Lots of data can leak if an IoT platform is left vulnerable and is attacked. Cloud-based IoT systems naturally require that data is shipped out to a cloud provider. Transmitting critical data between two entities extends security concerns from a company’s site to the third-party cloud provider and to any networks in between. IoT platforms that utilize edge computing solutions allow a company to maintain control of their systems and data and to eliminate multi-entity security concerns. Therefore, utilizing edge computing solutions can be advantageous from a security perspective.  
• Connectivity: Network reliability, stability, and speed are critical for IoT platforms. Companies can’t risk – even on the off-chance – that their IoT systems go down due to a significant network or cloud outage. Therefore, edge computing solutions are the right call from a connectivity point of view.  
• Scalability: System scalability has become very important as the number of IoT devices continues to expand. Cloud-based IoT platforms that store and process data in bulk in the cloud can be advantageous where scalability is key.

Cloud computing and edge computing are often compared and contrasted. However, since both cloud and edge offer unique advantages and disadvantages, there is no clear winner or loser of this cloud versus edge debate. While each approach can certainly be effective on its own, the best IoT system design may very well utilize a “best of both worlds” hybrid approach that deploys both edge and cloud solutions in concert. A hybrid system architecture can provide significant benefits by utilizing the combined strengths of both computing approaches.