There's a growing trend for the need for edge technologies when implementing an Industrial Internet of Things (IIoT) solution. Here I'll explore how the impact of edge can be increased through the adoption of edge analytics models. Taking it beyond data acquisition to deliver a solution where insights are democratized across your entire organization, regardless of location.
What’s the edge?
Firstly, let's define what we mean by the edge. Edge is one of the latest focus points within IIoT solutions. It's a key element to ensure that data at the edge is available.
Industrial organizations, or those that rely on industrial equipment (pumps, compressors, heat exchangers, boilers, etc.), often face a challenge when they want to adopt a digital transformation strategy including IIoT (or Industry 4.0). The reason is because the equipment or asset may be situated in remote locations, possibly even moving, for example, a container ship.
There are a wealth of savings and insights from these remote assets which can be used to:
- improve operations
- reduce costs
- increase equipment availability and productivity, and
- improve employee safety
However, there's often a challenge to collect data generated by these equipment and provide the analysis that modern data science techniques allow.
Edge computing solutions are evolving and allow the data to be collected and transferred to central repositories, typically cloud-based. Once gathered, you can then apply machine learning algorithms (models) to provide remote insight and allow you to:
- understand if your equipment is being operated optimally
- monitor key performance indicators
- provide analysis on important key resources i.e. fuel or emissions
- identify and replicate benchmarking and best practices to provide insight into health and safety situations
Edge analytics - Taking the next step
Going back to our ship example, it's a giant leap to now have the data analytics delivering insights to the central operations and business teams. But what about the crew on the ship's bridge? Delivering insight at the point of data generation; at the asset; may be where the insights are most valuable, giving immediate insights to the team actually operating the asset. This may provide the greatest benefit for them and allow them to take immediate action.
This is where the latest development in edge computing comes into play - edge analytics. This is where the exact same models being generated to operate on the cloud-based data can also be deployed back to the edge devices. Analytics models are developed by the data science team and paired to real-time data streaming from the asset. With the right system architecture in place, these models can be published back to the remote edge device.
In this way, we can impart immediate insight to the asset operations team regardless of the consistency of asset connectivity. It's rare to find a ship with a constant connection but with the models executing on the edge device we can provide the same level of insight across the organization. This is how we really start to take advantage of edge technologies.