What are the most effective MES deployment strategies for Industry 4.0?

While some Manufacturing IT leaders may still have skepticism about cloud-based MES, and in some cases, rightly so, due to connectivity issues and the mission-critical nature of MES in factories. Yet, cloud-based MES aren't new. Traditionally, enterprise-scale manufacturers have always centralized infrastructure and hosted MES in distant data centers from their manufacturing facilities to consolidate operations and leverage economies of scale. With increased access to the public cloud, they can use a lot more computing, storage, and networking options to suit their needs and leverage the benefits of the public cloud, such as elasticity, innovation, costs, and more, while also modernizing MES and integrating with emerging technologies.

Um ponto crucial sobre utiliza??o da nuvem tanto para implementa??o do MES como outras tecnologias, é a quest?o da seguran?a da informa??o, quando falamos de indústria 4.0 n?o podemos esquecer que para estarmos dentro deste conceito a conex?o fora da empresa é um desafio. A cadeia de suprimento conectada é segura é um ponto a analisar.

For a small company this is perfect a Cloud-based MES helps you to store the information you need to do the work orders, plus you can access remotely to add information you save time and money using this type of MES.

Something I have find during my recent experience is that the customer expectations must be clear, discussed and documented. Only after clearing out expectations a project plan must be revised to secure the results. This means that as a provider you should always keep in mind that quality starts in the voice of the customer. In other perspective, more technical is that previous requirements should be met, such as data standardization , material routings and process flows documentation is fundamental for a MES deployment.

Cloud based ensures standardised master data across plants, and benchmarking - but does have some drawbacks. Cloud based works well for enterprises with few plants, where the plants are geographically close to each other. Resilience and latency issues can occur if you have many plants in your landscape which are spread out globally. If the MES is mission critical be careful about choosing a cloud system. Cloud systems also don't work well when data must be kept in a specific country. Larger enterprises, or companies with mission critical MES would be better off with a hybrid solution.

Considering data security, Edge-based MES keeps sensitive data within the premises, reducing the risk of external breaches and ensuring compliance with data privacy regulations. This is important for industries dealing with proprietary or confidential information. Considering reliability for industries where downtime can result in significant losses, we need the edge devices which operate independently of internet connectivity, ensuring continuous operation even in environments with poor network conditions. Considering delays in data transmission to a cloud server could cause inefficiencies/disruptions in time-sensitive manufacturing processes. Edge-based MES processes data locally, minimizing latency and enabling real-time decision-making

Modern Edge computing isn't same as on-premise hosting. In fact, with the latest hardware and native compatibility with cloud technology, Edge is just considered an extension of the cloud to bring more benefits to customers, such as reduced latency. Some functions, such as quality inspection of high-volume manufacturing, safety-based use cases, and real-time data analytics and visualization, may need to be independent of connectivity, and sometimes, data residency regulatory requirements may dictate local processing. With modern Edge-computing, there can be rare instances where MES exists in an air-gap environment with no connectivity, either due to the operation's regulatory or secretive nature or physical connectivity challenges.

Edge based prevents global standardisation of procedures and processes. It's not possible to perform global benchmark and reporting without pushing the data into another system. This is the old model, some manufacturers still choose this, but the trend is to move towards more hybrid models in more advanced systems.

Hybrid MES takes advantage of both worlds. You can decide on Edge and Cloud distribution based on the earlier criteria. In cases like inference from machine learning models, you can build and train the model in the cloud, taking the benefits of high-performance and low-cost computing in the cloud since building and training are independent of latency requirements. Yet, once the model is built, you can deploy it on Edge, as inference might be latency-sensitive. Similarly, you can process real-time time analytics at Edge and more advanced and long-range analytics in the cloud, combining MES data with other data sources. Hybrid MES allows standardized workloads in all plants by simultaneously allowing remote updates.

This is the best of both worlds, but only a small number of MES applications do this justice (Infor MES being one of the most advanced). Many MES applications offer a hybrid model, but with very limited edge functionality - so be careful.

Before making the decision on cloud or edge based MES think about your network infrastructure. Keep in mind the fallacies of distributed computing and see which ones will trip you up. Maybe your network is unreliable, or you have limited bandwidth. All things to consider when making the decision. Additionally make sure you have a plan B which does not require a lot of manual intervention. If the MES goes down the system should respond automatically. Operations should not grind to a halt but continue, maybe in a reduced manner but production should not stop.