HARTING Deploys Edge Computing in Its Own Production

Summary

Industry users have debated the merits of initiatives like Industrie 4.0 and the Industrial Internet of Things (IIoT) for many years. Now, the wait-and-see period is over. As practical IIoT products and solutions emerge, users are starting to implement them to learn how they can create value. But with all the buzz about IIoT, where does an industrial engineer start? At the edge, of course!

The market for edge computers is growing quickly. HARTING, a supplier of automation components and solutions, uses its own edge computer products in its own plants to optimize injection molding machines. The experience gained has resulted in a growing suite of available applications based on the company’s experience with its MICA edge computer.

Edge Computing and Its Impact on Manufacturing

In network jargon, the “edge” refers to the locus of all endpoints of the extended internet. Edge devices can be computers, controllers, or any smart devices that sit at an endpoint with no further devices connected “behind” them. In the manufacturing world, edge devices include PLCs and industrial PCs, but can also be smart, connected sensors and actuators that share their information with their networked peers.

In manufacturing, “edge computing” typically involves the use of small, yet robust computers installed on the machine or close to the process. In a perfect world, devices of all sizes will be smart and connected and information shared seamlessly according to accepted standards. While we aren’t quite there yet, edge computers can help fill this gap. Edge computers typically are “headless” PCs packaged in a robust enclosure that can be retrofitted to existing machinery to quickly add functionality. The range of potential applications is broad, but many users today are starting out by deploying these devices to collect, analyze, and share data above and beyond what was possible using traditional PLCs and industrial PCs.

Edge computers typically work autonomously to perform a specific task, such as gathering and storing data from sensors and other devices and then evaluating the data and acting on the results. An edge computer may operate independently of the automation architecture, or connect to and share data with devices like a PLC or an industrial PC.  Some edge computers are called “gateways” because they connect devices that use different communications protocols.

HARTING Goes to the Edge

HARTING is a traditional supplier of connectors and other electrical components to a wide range of industries. In 2016, the company recognized the trend to edge computing in manufacturing and introduced a new line of devices under the “MICA” (modular industry computing architecture) brand.  

The MICA product line features ruggedized, energy-efficient edge computers with modular hardware that run open-source Linux-based software. Using virtualized Linux containers, multiple applications can run simultaneously on MICA, isolated from each other. The modular hardware and adaptable front panels allow function-specific boards to be developed and mounted in the same IP67 enclosure. Finally, the Linux OS allows use of a wide variety of development tools and programming languages.

MICA in Manufacturing:  Learning by Deploying

HARTING manufactures more than ten million individual parts every year in six plants around the world. To produce all those parts, the company maintains a fleet of 80 plastic injection molding machines from various manufacturers. Like any manufacturer in a competitive market, HARTING’s engineers strive to lower production costs by continuously improving manufacturing processes. According to the company, these challenges include traditional targets like reducing stock, shortening changeover times, speeding up delivery, and producing in smaller lot sizes.  In addition, the company also identifies challenges specifically related to injection molding machines. These range from dealing with machines from different OEMs, to individually optimizing performance and extending the useful life of each machine.

To tackle these challenges, HARTING set out to retrofit its injection molding machines with its own MICA edge computers running several different applications. One application connects legacy machines with a protocol translator programmed in a MICA. The program remaps information sent in the old EUROMAP 15 protocol from the early 1990s to OPC UA or MQTT messaging so the information can be read by other machines.

In another application, MICAs collect and organize process data from the machines and generate HTML-based dashboards to easily visualize the data with any internet browser. In a similar application, a MICA is configured to collect data from a variety of condition monitoring sensors, including air pressure, oil detection, and flow rates. These applications expose previously hidden data, making them transparent and therefore easy to monitor.

The data exposed by the MICAs enables plant personnel to analyze condition monitoring information and find the optical operating point for each machine based on the results of the analysis. Using analytics software, operators can define thresholds and set alarms that trigger when limits are exceeded. Trend data are plotted for visual analysis of historic data. A wireless connection also allows dashboard data to be displayed on smartphones and other mobile devices. Armed with these tools, HARTING has implemented predictive maintenance to help operators decide the best time to maintain each machine, rather than relying on fixed maintenance schedules.

No IIoT initiative is complete without a calculation of the return on investment (ROI).  According to HARTING, an initial upfront investment of €35,000 was made to procure and program the MICA edge computers. HARTING claims that the initial savings in the first year amounted to €41,000 in time savings and error detection. That translates to an RIO of just over ten months. The company further claims annual savings of just over €103,000 resulting from reduced time and effort to configure and reconfigure machines (now performed automatically rather than hand-carrying new configuration cards to machines on the plant floor).

Other successful applications of MICA include an automotive supplier that implemented a condition monitoring application for an existing overhead conveyor system that transports car chassis during assembly. The goal of the application was to monitor low-frequency vibration in mechanical parts of each fixture such as wheels, axles, and bearings. When wear and tear are detected, the fixture is taken out of operation for repair and replaced temporarily with a reserve fixture. Developed with the help of a system integrator, the user expects the application to extend the life of the overhead conveyor system by ten years.

Looking Ahead:  How Will the Market for Edge Computing Develop?

While nascent, the market for edge computing in manufacturing is developing quickly. As end users and machine builders recognize the value-add of such devices, the market will experience a phase of high growth, driven by the ever-growing variety of applications. Currently, edge computers are popular with end users that bolt them onto existing machinery. However, we expect machine builders to embrace the technology as well by offering both added functionality for new machines or retrofits for existing equipment.

MICA’s Solution Roadmap

In a market of generic devices, HARTING has recognized the need to differentiate its product by tying it to applications developed internally and by partners. Solution packages developed for target applications (i.e., injection molding machines) can help reduce integration costs and share application know-how with users. HARTING’s vision is to build a global net-work of partners that sell packaged solutions running on MICA hardware. 

To support the growing community of MICA users, HARTING has founded an open user group called MICA.network. Modeled after other industry consortia, MICA.network fosters information sharing among users to support and grow the ecosystem around MICA. In addition to end users, members also include technology providers that supply components compatible with MICA’s modular hardware, and solution providers that offer complete applications designed around MICA. Examples of available solution packages include applications for injection molding, condition monitoring of overhead conveyors, and asset management for machine tools.

Recommendations

Based on ARC research and analysis, we offer the following recommendations.

• For end users: Edge computing may be new to the world of industrial machinery, but plenty of solutions are already available. These devices can help optimize operation of existing machinery by exposing previously hidden data. They also typically offer a fast return on investment.
• For machine builders: Machine OEMs can develop and retrofit edge computer-based applications to existing machinery in the field. These apps add new functionality for local monitoring or serve as gateways for remote monitoring by the OEM, opening new revenue streams for digital services.
• For solution providers: Edge computers are an ideal, robust platform for packaged solutions in a wide variety of machine applications.  These range from condition monitoring to support for existing and legacy communications protocols.

ARC Advisory Group clients can view the complete report at ARC Main Client Portal or at ARC Office 365 Client Portal

If you would like to buy this report or obtain information about how to become a client, please Contact Us     

Keywords: Edge Computing, Gateway, Industrie 4.0, IIoT, MICA, ARC Advisory Group.