The next frontier for IoT: Edge analytics
As any technological system grows, it can either be advanced or surpassed by other technologies. Sometimes, this lesson is learned the hard way. For instance, many legacy photography companies either pivoted or shuttered with the advent of cloud photo sharing, mobile phone cameras and the improvement of home printers.
Even the most buzzed about technologies of the past decade are seeing disruption rear its ugly head — including IoT, cloud computing, social media platforms, virtual reality and the like.
With the internet of things, we are at an inflection point. Data being collected from devices, buildings and sensors is vast, so massive that, a lot of the time, it never is used because the data isn’t understood. Part of the problem with IoT data — and more specifically, industrial IoT data — is that it is collected on central servers and not the IoT devices themselves. Most of the time, these servers are either on the cloud or part of an in-house data system.
For a manufacturer in need of real-time IIoT data analysis of a device or sensor, this creates added complexity and an unnecessary blocker to getting information directly from the system. If you need to retrieve data from a source other than the IoT device built into production, you may not know where the data is specifically coming from, or worse, you may not be able to connect to the device and its related information to make integral business decisions.
The solution is bringing the analytics to the edge, which allows data to be analyzed at the point where the equipment is actually transmitting the information and there is no networked cloud or server data to sift through.
Edge analytics is perfect for manufacturers who need to be able to analyze and take the corresponding action in response to the massive amounts of data transmitted by IIoT sensors or the data transmitted from the production line. Beyond cutting down reaction time and the sifting of vital data, edge analytics also increases data security, especially in production facilities that create a continuous stream of data ripe for data attacks.
Here are a few more benefits for manufacturers who implement edge analytics processes into their IIoT programs:
Varied connectivity and data mobility
Implementing edge technologies removes the potential downtime risks and connectivity issues often inherent in production lines and manufacturing centers. Edge analytics systems can operate in places that might limit or require intermittent connectivity to the cloud.
Instead of relying on access to networks or the cloud for computing, storage, backup and analytics in manufacturing facilities where the infrastructure is often weak, businesses can have more faith in their sensors or devices processing and collecting operational data than if tied to servers or the cloud.
Need for real-time decision-making
In manufacturing, decisions need to be made as quickly as possible. Additionally, any problems or complications to production lines or automated process need to be identified and managed as quickly as possible.
Edge analytics allows data to be processed instantly, at least in sub-second speeds. For technologies like advanced robotics or automatic manufacturing line machinery, for example, the quicker issues can be identified and data can be analyzed is integral to the business. IIoT devices and sensors need to be able to do analytics locally without first sending data to the cloud, so decisions can be made rapidly.
Localized compute power
Many IIoT sensors and devices have space constraints due to the nature of manufacturing. Edge analytics hardware is lightweight and rugged, which is ideal for production lines, warehouses and other manufacturing needs.
More than anything, this creates an environment in which fast, secure and confident decisions can be made at the device level without the support of bigger computing power. This ensures reliability and uptime performance.
New storage and security needs
If you peruse any cybersecurity industry journal or dedicated blog, you will see how recent technological advancements — IoT and the cloud, to name a couple — are prized by hackers with malicious. The reason being that all these nascent technologies have easily exploitable loopholes that have not yet been solved by the market and the industry security systems.
As the numbers of sensors and IIoT manufacturing devices generating data on remote and sometimes mobile devices grow, so does the need for not just efficient storage, but data that can be secured in a variety of environments. Limiting the transfer of data to one step versus moving to servers or the cloud eliminates an easily exploitable threat.
Edge analytics is not a fad to be ignored
By not exploring edge analytics, manufacturers are limiting the potential benefits they can reap from their IIoT platforms. When the transmission of data, the ability to quickly analyze data and information security are not localized to a device or product in the manufacturing industry becoming more reliant on IIoT, then the potential for disastrous lags in response time or the loss of key data analytics occur without any even noticing.
Limiting where IIoT data analysis can be completed affects not only what manufacturers can offer to their customers, but also the amount of value that can be derived by advancements such as an IIoT-enabled production operation.
How edge analytics improves the system
To understand how important edge analytics could be to manufacturing, imagine if you will, a remote monitoring operation of the paint spraying process on an automotive line. The amount of paint applied in this specific process is critical to quality. The under- or over-application of paint is unacceptable and requires significant reworking and sunken costs.
When the performance data of this process is analyzed by IIoT devices and is then relayed across the cloud, the ability to respond with the necessary amount of speed to detect these fluctuations and the take action is lost. Facilities undertaking this car production process also often lack proper connectivity, which makes data transmission beyond the location of the device unreliable. As such, fluctuations in the painting process can be missed or detected too late to elicit a response.
To overcome these issues, edge analytics can be employed at the local level, where the IIoT devices are operating, allowing for the measurement of specific, combined operating parameters and issue alerts when these complex parameters are breached.
Building edge analytics into manufacturing is similar to starting an IoT system; you need to start with a simple threshold alerting systems that are easily and quickly understood by production engineers, product managers or field service technicians. This allows for a shift from a proactive approach to IIoT to a truly predictive data collection and analysis model.
All IoT Agenda network contributors are responsible for the content and accuracy of their posts. Opinions are of the writers and do not necessarily convey the thoughts of IoT Agenda.