While most organizations are prioritizing DX efforts, few have fully leveraged cloud and artificial intelligence (AI) technologies so essential to DX operational efficiencies. Now, that could be changing.

By Lisa Eitel | Executive editor

---

Digital transformation (DX) initiatives are at the forefront of business initiatives around the world. Spurred by rapid adaptations instituted during the COVID-19 pandemic, DX spending could reach $3.4T and add $100T to the world economy over the next two years, according to MarketsandMarkets.

While most organizations are prioritizing DX efforts, less than half have increased use of cloud and artificial intelligence (AI) technologies so essential to DX operational efficiencies. It’s a slow-burn existential threat to some — especially those organizations yet to embrace newer workplace collaboration and productivity tools. Even companies that have embraced DX initiatives and benefitted from their efficiencies often fall short of progress metrics, as programs often need continual fine-tuning, employee consent, and change management.

Manufacturing and automation have somewhat higher adoption and success rates than global averages of all industries. So, to learn more about the specifics of how DX technologies are being instituted in these fields, we recently asked several industry experts about trends they’re seeing in this space. Here’s what those experts had to say.

MEET THE EXPERTS
Chris Caldwell | Product manager – material handling • Yaskawa Motoman
Eberhard Klotz, Dipl. Ing. MBA | Global sales director — Industry 4.0 and digitalization • Festo
Dusty Schafer | Manager — software engineering • Kollmorgen
Frank Mignano | Sales manager — condition monitoring • Schaeffler North America
Gian Sachdev | Marketing head – Americas demand generation • Cognex
Josh Leath | Senior product manager — thermal • Yaskawa Motoman
Kurt Ledoux | Business director — medium voltage drives • Yaskawa
Michal Wierzchowski | VP of operations • Jabil
Mike Korkowski | Operations manager • LinMot USA
Richard Halstead | President • Empire Magnetics Inc.
Robert Cachro | Program manager — growth and innovation • Dynapar
Stuart Graham | Business Development Specialist • HEIDENHAIN Corp.
Thomas Burke | Global strategic advisor • CC Link Partner Association (CLPA)

How do you see automation making more use of data?

Graham: We’ve seen an increase of data collection to support IIoT and Industry 4.0 initiatives — and with that a need to efficiently store and analyze information. Such information could be that from simple temperature readings throughout a machine to inform a user about when that machine is operating at its peak efficiency. We also see a continual push towards predictive maintenance to identify when machine servicing is really needed … and help avoid failures.

Klotz: Predictive-maintenance, predictive-quality, and predictive energy programs are beneficial to production. Many Festo end users (especially small to medium-sized enterprises or SMEs) start with on-edge projects involving a few machines. Here, standardization of data is on protocols such as MQTT or OPC-UA. In contrast, most large endusers already have on-premises data lakes [centralized secured repositories] or even cloud-based their IoT platforms on which Festo can implement major projects.

Mignano: Connectivity has continued to improve over the past year. The specific task of machinery monitoring starts with edge computing — sensors deployed throughout a manufacturing facility. These communicate via protocols such as BLE, BLE MESH, LoRa, and 900 MHz to relay information about the plant assets such as pumps, motors, gearboxes, fans, compressors that keep manufacturing processes functioning. Some sensors even connect to cellular, Wi-Fi, and LAN-networked gateways that periodically send data to a unique cloud tenant.

If desired, collected information can then go to onsite data historians or computerized maintenance management systems (CMMs) to get actionable information.

If the data makes its way to the cloud, possibilities abound. Collected data can populate mobile devices and computers via email and text; AI processing can convert raw data into actionable information. Plus, from the cloud, all this data and actionable information is easily integrated into operator-inspection work-process applications via standard interfaces such a REST API, for example.

Wierzchowski: In shaping the future of Jabil’s manufacturing, we have achieved data normalization. This means our machines now stream data that is standardized and recognizable across the entire enterprise to support direct comparisons, analyses, and aggregation sans superfluous translation layers. We have accomplished this by establishing a uniform connectivity infrastructure for all our machines, forming the foundational layer of our manufacturing ecosystem.

What interfaces are supporting today’s data transfers?

Korkowski: LinMot automation components are compatible with all the most common fieldbuses. We have a new multi-interface drive that lets endusers choose the fieldbus they want when it’s installed. This reduces waste and inventory by having a drive that can work with many different fieldbuses allowing a future proof design.

Klotz: For years, Festo has supported the use of fiber-optic connectors. For example, several of our valve terminals destined for harsh environments feature fiber-optic connectivity.

Graham: Interfaces at the component level that transmit complex data (beyond basic information provided) go a long way towards simplifying machine designs. Instead of several I/Os with cabling and wires from every direction coming into the controls, things can be streamlined to localized data hubs and communication channel back to the controls. This in turn trims cost and sets up the overall architecture for success.

What’s emerging for fog computing and distributed control?

Klotz: Festo’s AX Industrial Apps are AI-based added-value apps that simplify predictive maintenance and predictive energy implementation. They’re also container-based, so can run on a container runtime on-edge (at an on-premises industrial PC) or in the cloud.

Cachro: The Dynapar HS35iQ encoder with PulseIQ technology is an addition to the field of smart manufacturing. Its capabilities include:

• System-fault indications for quicker troubleshooting
• Programmable output for increased flexibility, and
• Advanced monitoring for detecting various issues such as cable integrity, coupling slip, and temperature fluctuations.

The encoder’s comprehensive software tool offers detailed insights for maintenance and operation. These features enhance machine performance by supporting the creation of interconnected and intelligent manufacturing ecosystems.

T. Burke: The latest CLPA protocol is CC-Link IE TSN. It leverages Ethernet with new updates for time-sensitive networking (TSN) functionality. That allows the combining of control networks with realtime and general communications into a single Ethernet. When combined with edge connectivity through gateways, it renders all devices — control-oriented as well as information-oriented — available through edge access. This in turn facilitates troubleshooting and performance management. Our partner Mitsubishi Electric Edge offers gateway modules to access PLCs and information — and share information with cloud resources.

Mignano: There’s been significantly increased availability of wireless vibration-monitoring solutions for rotary bearings and rotating machinery. Such technologies have been made possible by the combination of enhanced battery capability (thanks to lithium thionyl chloride-based systems) and low-power micro-electromechanical systems (MEMS) sensors.

The advancement of these two technologies in robust packaging and advanced communication protocols has enabled myriad wireless battery-powered sensors to arrive on the market over the last three to five years. In contrast with past systems (from more than five years ago) today’s wireless vibration-monitoring solutions are robust, cost effective, and functional. They also provide ROI in months or even weeks in some cases.

Sensors that can measure vibration as well as temperature, pressure, flow, humidity, and so on are now available from various vendors for a broad range of applications. Edge-computing devices and data availability are now common and integral to the reliability strategies of many leading organizations. This data is also nurturing emerging and existing AI and machine-learning technologies from many of the same vendors. The result is improved asset reliability, less unplanned downtime, and greater profit margins for manufacturers.

Klotz: With our edge devices, Festo runs AI projects for end users without their own hardware. They devices are built on extended PLC platforms or IPCs, have 16 or 32 GB of storage, support MQTT and OPC-UA data formats as standard, and offer IT connectivity.

Caldwell: First, more automation solutions today incorporate smart sensors and cell-level event tracking. These support intuitive root-cause analysis and realtime KPI analysis and visualization. Second, systems such as Yaskawa robots can couple with Yaskawa Cockpit — a tool to gather realtime data from robots, drives, servomotors, and nearly all OPC-UA devices installed within a enduser’s operational environment. Cockpit allows for data collection and visualization on a local edge server.

What’s the latest in industrial cybersecurity?

Ledoux: There’s increased demand for us as a supplier to provide communications (such as monitoring and control over the network) with increased cybersecurity. Our equipment resides inside the customer network, and security is typically the responsibility of the customer and their IT department. So, the question is this: Do we need to consider other methods inside existing firewalled networks for a secondary form of security?

Caldwell: As industrial-cybersecurity risks become more present for all sizes of companies, there’s a push to supply and support solutions based on local edge servers. If end users need cloud-based solutions integrated into their operations, local IT staff can always securely extend an edge server’s collected information into their cloud service.

Halstead: With the proliferation of hackers, bank losses to data thieves, mounting ransom attacks, and the U.S. government’s push to protect data, there are lots of resources getting thrown at cybersecurity. As obvious money targets get hardened, are going to turn their talents to more indirect ways to extract funds. For example, hackers might turn off computer-controlled equipment to shut off a city’s water supply and then demand a ransom to release the equipment. The list of possible approaches is very long.

Protection solutions all cost money. Typical industrial equipment provides no form of hacker prevention or data encryption. Less costly than changing the operating equipment is to add hardware that acts as a firewall requiring authentication to access a facility’s equipment. As we’ve learned from TV shows, good hackers can (with enough effort) bypass such protection. But if equipment makes the problem difficult enough, hackers move on to easier targets.

Schafer: Awareness and education about cybersecurity risks are increasing. The topic is confusing, as there are multiple standards, various horror stories in the news media, and an array of companies that sell cybersecurity products and services. Sometimes advanced technical solutions are necessary, but many times simple best practices are the first line of defense.

The Secure by Design trend sponsored by the U.S. government’s Cybersecurity and Infrastructure Security Agency (CISA) is an important industry change. The idea is to incorporate security measures by default instead of adding them afterward. This is a superior methodology to secure environments and improve customer experiences. It doesn’t fully address issues with existing or legacy products, but it’s a step in the right direction.

Klotz: Cybersecurity is increasingly important in general. Governments all over the world define regulations with which entities must adhere — for example, in the EU through the Cyber Resilience Act or CRA. These define measures and ways to boost the cybersecurity of industrial communications and software. Sometimes, satisfying standards such as IEC 62443 are a good way to ensure satisfaction of upcoming cybersecurity laws.

Key is to establish necessary processes and roles during product development and product lifecycle management — and to monitor for new vulnerabilities and issue security patches (updates) when needed. Festo is active in this.

Some industry segments (including many process industries) define their own cybersecurity requirements on the field level — as in NAMUR NOA definitions or advanced bus communications such as the Ethernet-APL. On the other hand, practitioners struggle when sharing computers or mobile devices and often prefer to have a single user for many operations.

Halstead: Organizations serving the defense industry are being required to heavily invest in new levels of security. The U.S. government has implemented requirements for cybersecurity that include physical security. They’ve basically made everything that the government buys a form of classified information, but with a different name — Controlled Unclassified Information or CUI. If a company doesn’t get with this program, in time the buyers won’t be able to use that company.

Sachdev: IIoT and Industry 4.0 have been buzzwords for years, but now the type of automated data-rich factory environment to which they refer is increasingly attainable all the time. For us, feeding image data into IIoT applications is a core function, as that data can fuel both application-specific process improvements and larger-scale enterprise resource planning (ERP) improvements. In fact, in our view, machine vision is essential to make IIoT a reality by connecting the IoT to the Internet.

We contribute to DX through the Cognex Edge Intelligence platform, which transforms big data into smart data and provides realtime system performance monitoring and device management to improve overall equipment throughput.

How is networked manufacturing changing installation approaches?

T. Burke: One of the most important criteria in modern equipment is the use of an Ethernet backbone to let all devices exist on a common cable. This promotes the access and management of devices for improved configuration, updates, and troubleshooting. Smarter devices such as machine vision can integrate with controllers and be managed with protocols such as SNMP. If devices are isolated to separate control networks, those management opportunities are unavailable.

Klotz: IoT-relevant field devices and machine PLCs must support data standards such as MQTT and OPC-UA — and time-synchronized clocks are also mandatory. Realtime decision making inside machines is better supported by edge devices and smart products that automatically adapt to processes. These decentral intelligent assets demand new forms of IoT management for releases, security, access roles, and so on. No wonder such added-value applications increasingly running on industry PCs offering container runtimes to simplify deployment.

Any interesting DX applications you’ve seen over the last year?

Klotz: One premium automotive supplier used a DX approach to address an issue involving welding guns. After a few proofs of concept (POCs) to fix the problem, the supplier rolled out the solution to all its global production sites. The result was 25% less unplanned downtime, and 20% faster mean time to respond (MTTR).

Elsewhere, a largest automotive brand installed a complete production line to standardize all data. This brand chose Festo as a turnkey partner for the AI analytics of all the brands associated with that production line. Upfront effort was quite high but promises high long-term savings.

Leath: There are a lot of data points one can collect from a robot. For example, in a robotic welding application, it’s possible to track arc on time, the amount of wire consumed, cycle times, production counts, errors, and more. This data can drive efficiency to optimize production processes and add traceability to parts. The latter ensures parts meet their original specifications.

We can also track who was logged into the robot if changes to a job are made. All those data points can be analyzed by tools for predictive maintenance — as when maintenance personnel are due to change out the grease on an axis, or if a motor shows signs of failure. After all, no one likes a surprise production stop.

More cutting-edge uses employ that same data plus additional information from additional sensors to teach AI models. That’s helpful for inspection or sorting, for example … but usually requires a great deal of data and processing typically done with a cloud-type application.

T. Burke: With PC-based HMIs, better communications for information management purposes are enabling a greater ability to perform supervisory control, data acquisition, and advanced analytics. We’re seeing both on-premises and cloud solutions for enhanced analytics; companies such as our partner ICONICS are valuable contributors to solutions from Mitsubishi Electric and others.

In fact, panel-based HMIs are also gaining in capabilities and benefiting from greater integration with the programming software used for system configuration. Engineers can configure the PLC while also automatically populating the data structures of the HMI — and deliver objects to match the control elements in a project. We’re also seeing a greater level of integration with SCADA and Cloud-based data repositories directly from panel HMIs.

Design World | designworldonline.com/trends

---

Filed Under: DIGITAL TRANSFORMATION (DX), Trends
Tagged With: festo
 

---