Skip to content

Edge AI increases productivity and efficiency

AI on the Edge is an emerging IT architecture that leverage the power of distributed computing to achieve higher efficiency, lower response time and cost savings.

Edge AI in practice

In practice, this means that AI processing, performed by Machine Learning algorithms, happens at the local level, where data is generated (for instance by machinery, IoT and IIoT – Industrial IoT – devices).

Bulk data, therefore, does not need to be transferred via the Internet to a central or cloud based data processing facility (usually a server cluster) and then wait for the result to be sent back.

This allows for critical and time sentire decisions to be taken in real-time, as the whole process takes milliseconds instead of seconds, eliminating the delay of back and forth data transmission (which is technically called “latency”).

In a centralized model, where all data processing happens in one (or few) locations, the bandwidth necessary for all data transfer represent both a bottleneck and a considerable expense. Edge AI solves both problems by completely eliminating the need for continuous bulk data transfers.

A technology enabler

Edge AI architectures are already more widespread than one might think, especially in the consumer world.

One such example is that of the ever-increasing popularity of self driving cars, the likes of Tesla. It goes without saying that the on-board artificial intelligence must continue to operate even in areas where there is bad or no Internet connection. In other words, the car must be able to analyze all data coming from on-board sensors and interpret it autonomously. Not only this applies to situations where there is spotty data connections, but the reaction time of the self-driving car must be as fast, if not faster, than that of a human driver. If the system had to rely on a centralized data processing facility, because of latency all decisions would take one or more seconds, a reaction time completely inadequate to real-world situations, which would endanger lives, thus impeding the development and use of self-driving cars altogether.

Edge AI in production and for manufacturing

The implementation of an AI on the Edge model not only benefits consumer products and applicationbut also brings significative advantage in the Industrial sectorboth solving complex problems and allowing for smoother and more efficient processes. 

This addresses and solves a series of challenges typically found in production plants:

Volume and variety of IoT data

Many factories utilize both modern and legacy manufacturing assets and devices from multiple vendors, with various protocols and data formats. Although the controllers and devices may be connected to an OT system, they are not usually connected in a way that they can easily share the data with IT systems as well. In order to enable connected manufacturing and emerging IoT use cases, a company needs a solution that can handle all types of diverse data structures and schemas from the edge, normalize the data and then share it with any type of data consumer, including Big Data applications

The complexity of real-time data

in order for a plant or a company to drive predictive analytics use cases, a data management platform needs to enable real-time analytics on streaming data.

The platform also needs to effectively ingest, store, and process the streaming data in real-time or near-real-time in order to instantly deliver insights and action.

Freeing data from independent silos

Specialized processes (innovation platforms, QMS, MES, and so on) within the manufacturing value chain reward disparate data sources and data management platforms, that tailor to unique siloed solutions. These niche solutions limit enterprise value, considering only a fraction of the insight cross-enterprise data can offer, while dividing the business and limiting collaboration opportunities. The right platform must have the ability to ingest, store, manage, analyze and process streaming data from all points in the value chain, combine it with Data Historians, ERP, MES and QMS sources, and leverage it into actionable insights. These insights will deliver dashboards, reports and predictive analytics that drive high-value manufacturing use cases.

Edge AI Advantages

  • Business critical decision can be taken autonomously and in real-time directly at plant level. 
  • It increases the level of security in terms of data breaches and data theft and it significantly reduces the privacy issue that might be associated with data as this no longer needs to be transmitted and shared in a centralized cloud or a centralized system. 
  • Along with a very significant reduction in latency, the reduced bandwidth needs eliminate the worries of unreliable data transfer speed and translate into a reduction in the costs of the contracted internet services.  
  • Contrary to centralized data processing systems, Edge technology devices do not generally require specialized maintenance by data scientists or AI developers. 

Success case

Industry 4.0-Manufacturing Edge Analytics for total productive maintenance

An Italian company, globally recognized as a best-in-class 
manufacturer of braking systems and components, with a strong focus on R&D and a keen interest in the potential of innovation and new technologies chose Agile Lab as a partner to adopt a new “Industry 4.0” solution enabling its TPM (Total Productive Maintenance) manufacturing best practices and innovating the Preventive Maintenance strategies. 

With a presence in 14 countries worldwide (and a workforce of over 10.000 employees) and the control of the entire production chain, including owning its own foundries, a decisive success factor proved to be the adoption of “EDGE Computing” concepts, computing done at or near the source of the data, running fewer processes in the cloud and moving those processes to local places, in the Plant itself 

solution enabled by Agile Lab’s Wasp, a feature-rich, highly scalable frameworkideal for an IoT/IIoT context and resulting in bidirectional coordination of Machine Learning modeling processes between HQ and the Plants increased both the efficiency of Data Scientists’ activities and the effectiveness of the AI models.
The i
mplementation of data flow analysis directly at the Plant Edge – without streaming high-frequency data to central HQ – allowed for a great effectiveness in predicting the correct lifetime of machinery assets, obtaining more reliable estimates of machinery RUL (Remaining Useful Life) and real-time alerting/predictive alerting capabilities. 

 

Fill the form on the right  to receive
a full detailed report of this success case in PDF form
directly in your e-mail.

 

Would you like to know more?

TALK TO OUR EXPERT