Realising remote monitoring goals

08 September 2019

Remote monitoring can offer a wide variety of benefits for industry, including a reduction in downtimes and an increase in productivity. However, there are some challenges that must first be overcome.

Having greater visibility and transparency of data that, traditionally, has been stored in decentralised silos makes it possible to address production priorities in a timely manner, based on reliable data. 

To achieve these remote monitoring goals, it is necessary to first overcome some challenges:

Challenge one – Connecting various systems causes network outage: As the IIoT continues to make inroads in manufacturing more and more machines and devices from different OT subsystems get connected to ensure transparency of shop-floor operations. In older hierarchy-based network architectures, these systems would run smoothly. However, when connected to each other, networks can become unstable or can fail due to slow responses, network interruption, or single point of failure, resulting in unreliable data and PLCs receiving incorrect responses. On closer inspection, these network problems are generally attributed to three issues:

• A cocktail-like topology: With regards to a cocktail-like topology, different machines and systems might use different commercial and industrial network devices as they were built at different stages. Subsequently, an OT environment contains many interfaces that may impact the daily operation of long-term hardware with regard to EMC, surge, and many more when the network gets connected.

• Unmanageable interoperability: Subsystems employing different OT protocols usually use unmanaged Ethernet switches to connect machines in a single system directly. In an interconnected scenario, each connected node in a port should not only be manageable, but it should also support different OT protocols to be visible in SCADA systems, thus ensuring interoperability and communication with each other in the network.

• Insufficient resilience: A resilient network is rooted in a network design that emphasises redundancy. A traditional star or daisy-chain network is easy to deploy and maintain; however, it cannot provide millisecond single-point-of-failure recovery when connecting multiple factory subsystems together. Another common issue is insufficient bandwidth, as it will make network connections unstable, especially multiple systems running on the same network. Bandwidth becomes inadequate as data transmissions increase in an   interconnected IIoT-based network.

One solution to simplify a network for variety of systems can be found in managed industrial Ethernet switches, such as those offered by Moxa, which provide networking solutions for cross-platform systems to transmit he data between plant-floor equipment and SCADA/MES systems or cloud platforms.

An industrial-grade design enables smooth operation in environments with a high number of interferences (e.g., EMI/EMC) and extreme temperatures. An automotive component supplier in China, for example, now uses Moxa’s industrial Ethernet switches, to help prevent packet loss, caused by a high number of environment interferences and extreme temperatures, which a previously used commercial-grade solution could not handle.

Moxa’s managed industrial Ethernet switches can seamlessly integrate devices to SCADA/HMI systems and support the mainstream industrial proprietary OT protocols, such as Modbus and EtherNet/IP. The network of one solutions provider of automated material handling systems in the USA, which used unmanaged switches to integrate its Ethernet/IP PLC with its existing SCADA system, experienced jitter as more and more devices were connected to the network. Moxa’s managed switch supports EtherNet/IP protocols to enable IGMP snooping, which controls multicast traffic so that jitter is eliminated.

Challenge two – Invisible large networks can be unreliable: Traditionally machines and other equipment involved in processes such as assembling, processing, and packaging were visible in SCADA systems but networking devices and other related equipment were invisible to operation managers. As a large variety of systems are deployed in different manufacturing sites on a large scale in the age of the IIoT and Industry 4.0, their invisibility is not a good thing in modern-day manufacturing. There is always the increased risk of network failure that can lead to catastrophic events during production. As networking equipment is an integral part of any smart manufacturing process, it is critical to make them visible and manageable, especially in large-scale applications. These challenges are vital to address:

Traditionally, different systems were cascaded into a large network to facilitate data collection and system management of machinery and equipment so networks only consisted of a ten or so devices. In the era of the IIoT, the same network might house hundreds or even thousands of devices. Inevitably, such large-scale deployments of devices come with network management challenges in managing a network, such as being updated in real time on the connection status of all the network devices.

As the complexity of the entire management of a network increases, pressure mounts on the network engineer to acquire new skills and tools to deal with new demands –for example, learning new network management software. Further, if IT teams manage the network, then it is most likely that their usual management tools do not support industrial switches, making their implementation difficult.

Operating multiple systems and software separately, managers need to use management software to see the status of the devices on the network. For factories operating 24 hours a day, dealing with network issues after hours or during business trips can be a huge headache as they cannot monitor the relevant data remotely. Here, an industrial network management solution can help.  MXview, from Moxa, for example, enables engineers to visualise the network to make monitoring and management of networking devices easy.

Case study
With its five-axis CNC machines already running 24 hours a day up to 7 days a week, Tech Manufacturing, a manufacturer of machine metal parts for the aerospace industry set out to increase the productivity and useful life of its existing machines through smarter operation and real-time performance data. “We needed a better understanding of how our machines were actually performing for us in real-time. Live and historical machine performance data would also help us identify technical or process issues that were detrimental to productivity,” said Jerry Halley, chief engineer at Tech Manufacturing. With a combination of hardware and software, a CNC monitoring system would collect, analyse, and visualise the necessary performance metrics. However, Halley needed to weigh the productivity gains of such a system against the cost and effort of deployment, especially if it involved a new and unfamiliar server-based IT infrastructure.The ideal system would be easily deployed without specialised IT equipment, knowledge, or effort, and would not require repeated software installation, updates, or configuration.

Each CNC machine was connected to the existing local area network, so no additional IT infrastructure was required. For legacy machines that did not have a readily available Ethernet port, Shop Floor Automations provided an easy-to-deploy solution that was developed with Moxa. With the local network connected to the Internet, machine performance data was easily viewed and analysed by cloud-based software. Key performance metrics were organised on a visual dashboard so owners and machine operators were able to see exactly how productive each cell was, down to the machine level.

With a cloud-based monitoring system, Tech Manufacturing is able to minimise upfront costs and deployment effort and, with cloud-based solutions a CNC monitoring system can be set up in less than a day, with almost zero additional IT infrastructure or maintenance effort.


Contact Details and Archive...

Related Articles...

Print this page | E-mail this page