Four technologies every modern manufacturer should adopt

Condition monitoring, vision system AI, message queuing telemetry transport (MQTT) and modern supervisory control and data acquisition (SCADA) systems can help manufacturers looking to improve their operations.

By Brandon Teachman March 26, 2024
Courtesy: CFE Media and Technology


Learning Objectives

  • Understand how plant floor data can change businesses for the better.
  • Learn about technologies that can provide new and better insights on short- and long-term operations.
  • Gain insight into practical artificial intelligence (AI) that can be implemented now.


Modern manufacturing insights

  • Utilizing sensors for equipment health assessment enhances production efficiency by predicting faults, reducing downtime, optimizing energy use, and enhancing environmental sustainability through targeted improvements.
  • AI-driven vision systems swiftly detect defects, streamline production flexibility, minimize waste, and aid quality control, offering adaptable, scalable solutions for various manufacturing processes.
  • MQTT’s efficient data sharing and modern SCADA systems offer universal connectivity, enhanced analytics, scalability, and IIoT readiness, revolutionizing plant monitoring and control for intelligent manufacturing.
  • The traditional methods are no longer sufficient to meet modern manufacturing demands, consumer expectations and environmental challenges. The world is moving too fast. Manufacturers must adopt smart manufacturing to increase production efficiency, quality, flexibility and sustainability.
  • What technologies are available to support smart manufacturing and how can they be implemented? Condition monitoring, AI-powered vision systems, message queuing telemetry transport (MQTT), and modern supervisory control and data acquisition (SCADA) systems are four key technologies that help turn plants into a smart factory.

1. Condition monitoring

Condition monitoring employs sensors to monitor the performance and condition of equipment. A condition monitoring system can identify wear, problem or failure indicators by gathering and evaluating data from these sensors before they result in significant issues. By doing so, companies can schedule maintenance and repairs in advance, saving money on downtime and increasing the equipment’s lifespan.

Condition monitoring can also improve energy use and reduce carbon footprint. Companies can find areas to save energy by monitoring components’ power usage and efficiency. For example, can an oven’s temperature be reduced when not in use or turn off motors when not needed? This type of monitoring of environmental factors also may provide better working conditions and can help reduce variance in the manufacturing process. Temperature, vibration, noise and moisture are critical for ensuring employee comfort and minimizing the negative impact on materials.

The sensors used for condition monitoring need a software platform to gather, store, process and display the data. There are application-specific software platforms and ones that are general. One of the most popular application-specific platforms revolves around moving parts like motors, pumps, fans and compressors. This software checks the vibration, temperature and more from a sensor mounted on the motor.

Some of these software packages also tie into the drive on the motor to get additional motor information to protect the asset better. The general systems can tie into multiple types of sensors and other devices from multiple manufacturers. The general systems require more setup, but have more flexibility. Both platforms can also link with other systems in the plant such as an enterprise resource planning (ERP) system, manufacturing execution system (MES) or computerized maintenance management system (CMMS). They also can use AI to predict future failures.

Four aspects to consider when selecting condition monitoring are:

  • What equipment has the most significant impact on manufacturing?

  • What data do you need to decide when to repair something proactively?

  • How do you want this data, and when?

  • How much time can you invest in setting up this system?

These criteria help narrow down what is the best fit. For example, if the only concerns is the uptime of four large grinders, some companies make a prepackaged solution that can be installed on large motors. This is a pre-made solution, quick to implement and has a fast return on investment (ROI) for an application.

Another example would be a plastic manufacturer looking to decrease downtime and quality. Based on preliminary data, they found temperature and humidity in the building were affecting their products. They also were having unplanned downtime due to motors failing. They went with a highly customizable general system that can be used with multiple manufacturers’ products.

2. Vision system AI

Machine vision systems give machines the ability to perceive and comprehend their surroundings. AI vision uses digital cameras and machine learning to analyze images and videos of the plant’s products and machines. It can detect defects, errors and anomalies in real time and adjust settings or alert team members. This allows the AI Vision system to help reduce waste, rework and downtime while increasing customer satisfaction and profitability without increasing labor. AI vision is also a scalable and adaptable solution that can be customized to diverse types of manufacturing plants and products. It can learn from new data and feedback and adjust to changing conditions and requirements. AI vision is a tool and a partner that can help the manufacturing plant achieve its quality goals.

Manufacturers are constantly seeking ways to increase their production flexibility and agility. By using vision system AI to recognize several types of products or components and adjust their settings accordingly, they can switch between different orders or batches without wasting time or resources. They also can use vision system AI to monitor their inventory levels and replenish them when needed.

When implementing an AI vision system, there is an investment in the correct cameras and lighting for an application. Companies also need a software platform that can provide pre-made or custom-made algorithms for specific tasks. The software platform also can integrate with other systems in the plant such as programmable logic controllers (PLCs), SCADA or MES to provide real-time feedback and control.

Five things to look at when deciding if AI vision is the right option include:

  • Are you making a repetitive product?

  • Do you have defects that are visible to the human eye or a thermal camera?

  • Can a defect be detected by a measurement check?

  • Do you have quality control (QC) documents with the variances you allow for the product?

  • Do you have a culture where automation is accepted?

One example of a successful implementation of vision system AI is the manufacture of consumer hand tools. During manufacturing, they use vision systems throughout the line to ensure the parts are in place, ensure components were correctly added and all required labeling is readable. This allows the machine to make corrections, if possible, without human interaction. When the machine cannot correct, the product is discharged from the line for manual interaction.

3. MQTT on the factory floor

MQTT is a communication protocol that permits data sharing between devices via the network. Devices can publish messages on topics and subscribe to topics that interest them under this publish/subscribe approach, which is the foundation of the system. MQTT is lightweight, quick, dependable, secure and scalable. It also helps that MQTT is already commonly found in modern devices.

MQTT can benefit manufacturers who want to achieve smart manufacturing by connecting their devices and systems housed locally or on the cloud. Using MQTT on the factory floor, they can easily collect data from their sensors, controllers and machines and send it to a central server or broker. They can also receive data from the server or broker and act on it accordingly. The systems interacting with this data can include an ERP, CMMS, MES, SCADA and more.

By standardizing on MQTT, a manufacturer can enjoy many benefits, such as:

  • Reduced network bandwidth and resource consumption: MQTT uses a publish/subscribe model, which means that devices only send and receive the data they need, avoiding unnecessary traffic and overhead.

  • Improved scalability and interoperability: MQTT can connect thousands of devices across different platforms and technologies, allowing for easy integration and expansion of the plant network.

  • Enhanced security and reliability: MQTT supports various encryption and authentication mechanisms, as well as quality of service levels, to ensure the data is securely transmitted.

  • Increased flexibility and agility: MQTT allows for dynamic discovery and configuration of devices, enabling the manufacturer to adapt to changing needs and requirements on the plant floor.

To get started with MQTT on the factory floor, there is little to no investment in equipment for most manufacturing facilities due to the wide adoption of MQTT. Companies also need a software platform that can act as a broker or server for MQTT messages.

Many brokers have free options to get started. To provide data visualization and analytics, these platforms may also interact with other systems in the plant including SCADA or MES systems.

4. Modern SCADA systems

SCADA systems allow manufacturers to monitor and control their processes and equipment. They can collect information from a range of hardware such as sensors, PLCs, VFDs and more. They can show it on graphical user interfaces (GUIs) like dashboards, charts and maps, among other things.

Modern SCADA systems are far more sophisticated than earlier versions. They also can provide manufacturers with more than simply monitoring and control options. Modern SCADA systems can provide these five benefits:

  1. Universal connectivity and IIoT readiness: A modern SCADA system can connect to various devices and data sources, such as sensors, controllers, databases and cloud services by using open standards and protocols, such as MQTT and OPC UA. This enables the SCADA system to leverage the IIoT benefits such as real-time data, analytics and remote access. Users also can link their equipment and processes to the cloud or enterprise resource planning (ERP) systems, allowing users to access data anytime, from any place and via any device. Companies also may employ the computing capacity of the cloud to do complex data processing and analysis using artificial intelligence or machine learning methods.

  2. Web deployment and mobility: Most modern SCADA systems are web-deployable, which means they can be accessed and operated from any device with a web browser, such as a smartphone, tablet or laptop. It provides flexibility and convenience for its users, as they can monitor and control their processes anywhere and anytime. Some modern SCADA systems also support mobile applications, providing customized and interactive user interfaces for different devices and roles.

  3. Data visualization and analytics: A modern SCADA system provides rich and intuitive data visualization and analytics capabilities to help users understand and optimize their processes. It also offers various charts, graphs, dashboards, reports and alarms that display the data in a clear and meaningful way. The data visualization is customizable and can show users just the data they want. The system can even be scheduled to email the data. Modern SCADA systems support advanced analytics, such as artificial intelligence and machine learning (AI/ML), which can provide insights and predictions based on the data.

  4. Scalable and modular: Most modern SCADA systems have been designed to be very scalable, which allows companies to add more equipment or manufacturing locations. The SCADA system comprises independent modules that can be added, removed or replaced. Each module has a specific function and communicates with other modules through standardized interfaces, allowing for flexibility and system customization. A SCADA system can distribute the control functions among multiple controllers or servers, which can be located at different sites or regions. This reduces the load and dependency on a single central controller and improves the availability and redundancy of the system.

  5. MES and HMI functionality: Modern SCADA systems can perform the roles of manufacturing execution systems (MES) or human-machine interfaces (HMIs), enabling manufacturers to manage their production planning, scheduling, execution and tracking in one system. In most cases, this reduces the complexities and system costs.

Where and how to start implementing

Harness the power of condition monitoring, AI vision systems, MQTT and modern SCADA systems to embrace the future of manufacturing. The first step is ensuring companies have the right culture in place. If the team is not open to change, implementing any of these technologies will be challenging. If this is not an issue, it’s best to start small. Select a smaller area or line with a known issue such as constant downtime, high rework rate or not hitting production numbers.

From there, determine how to collect data from that line and display it to the larger team. To start, this collection may be manual. Companies can then use this data to select the next steps. Once companies have the data, involving outside parties like system integrators and vendors is easy. This is because they have measurable data to work with.

If there is unplanned downtime due to motor issues, condition monitoring may be the solution to plan the downtime, but it also may provide data to find a solution to go longer between failures. Users also may find that the defects from this machine may be due to getting out-of-spec parts from downstream, and a vision system may do a better job finding this issue before more work is done to the part. It also may come down to having an issue that only occurs when someone is at lunch and therefore becomes a training issue.

Modern and intelligent manufacturing for the future

Intelligent manufacturing is here right now; it’s not science fiction. Condition monitoring, vision system AI, MQTT and modern SCADA systems are four essential technologies that can assist in making it a reality. These technologies can work together or independently to improve a facility’s production quality, flexibility and sustainability.

Author Bio: Brandon Teachman is an application engineer at Vision Control & Automation, where he helps businesses improve their manufacturing processes through automation solutions.