Background Image

Edge-cloud hybrid computing :

High performance in production



Data is central to improving manufacturing processes. This is because in order to improve a process, manufacturing managers first need to understand the current state of the process in quantitative terms e.g. processing time, cycle time, units produced per hour. Once the current state is well defined, improvement plans can then be initiated and the outcomes measured against the current state to indicate if the improvement was realized.

With the rise of cloud computing it has become easier for manufacturers to store and process large amounts of production data without the need to manage a complex IT infrastructure. Cloud computing allows manufacturers to pay only for cloud services they use, helping them lower their operating costs, run their infrastructure more efficiently, and scale as their business needs change. However, a pure cloud computing system poses significant challenges in a production environment.

A far superior approach is to use a hybrid edge-cloud architecture for the processing and storage of data. This hybrid architecture combines the best of both worlds while minimizing the problems associated with each architecture as a stand-alone option.

What is edge computing

Edge computing is an emerging computing paradigm which refers to a range of networks and devices at or near the user. Edge is about processing data closer to where it’s being generated, enabling processing at greater speeds and volumes, leading to greater action-led results in real time.

Challenges with cloud computing in manufacturing

Manufacturing is a high paced dynamic environment where it’s very important for operations personnel to get insights into their operations with minimal delay. One of the major challenges in manufacturing is the availability of high bandwidth internet. As the cloud relies heavily on high bandwidth internet, low bandwidth internet can cause loss of information being transmitted to the cloud from the production plant or the information is transmitted with high latency.
This causes the insights generated via the cloud to be either inaccurate or too slow for taking any meaningful action. Furthermore, processing large amounts of data on the cloud can be very expensive and therefore cost prohibitive for manufacturers. Depending on how much data is being processed the cost could be well over $100,000 per month per production line.

Benefits of edge computing in manufacturing

Edge computing reduces reliance on internet by processing data locally on a local server or edge device. This way the insights can be generated without any data loss and operations personnel can get access to the data quickly since it doesn’t have to be routed through the cloud. Thus production operators can use the data to take action and fix their processes in a timely manner. Processing data on the edge is also much cheaper than the cloud since the only cost is the cost of the hardware that can be depreciated over multiple years versus a continuous recurring cost on the cloud. The hardware in most cases can also support multiple production lines which further reduces the overall cost of the system.

Challenges with edge computing in manufacturing

While edge computing can process data a lot faster and cheaper than cloud computing, storage of data can become problematic over the long run. In a manufacturing environment, data is being generated every millisecond and it needs to be stored somewhere since it needs to be used for analysis. In a pure edge computing architecture, the data would have to be stored on a local server which would constantly need to be scaled up and maintained by internal IT teams. This can be very cost prohibitive for organizations with lean IT teams. The cloud is a cheaper and more reliable option for big data storage.

How i-5O’s vision system uses an edge-cloud hybrid model for optimal performance

At i-5O we’ve taken a best of both worlds approach by designing an architecture that takes advantage of the edge and the cloud. We deploy an edge device locally that runs our deep learning models with data post-processing but use the cloud for big data storage so that we don’t need to constantly grow the storage of local devices. Using this approach we are able to get insights to our clients with minimal latency and low data processing costs while minimizing data storage costs. This way our clients can easily access multiple years of data without any major increase in cost while maintaining high system reliability.

Try it for Yourself

Schedule your 1st Free Consultation and Start your AI Journey Today