How IIoT happened at Hirotec
Tactical engagements support a strategy to make better use of data.
Tier-one automotive supplier Hirotec recently completed a successful test-bed application project that exploits the latest technology advances in industrial connectivity and analytics.
The newly implemented system melds operations and maintenance data, and by doing so, it can help employees identify trends that lead to contextualized, actionable insights. Moreover, Hirotec managers said they are building a plant-floor technology infrastructure aimed towards the future.
"The idea of making use of the Industrial Internet of Things [IIoT] as a platform arose organically, but then it had to be ‘sold’ to internal stakeholders," Justin Hester, research and development project manager, said.
A supplier of car enclosures and exhaust systems, the Hirotec Group is a $1.6 billion corporation with 26 facilities in nine countries. It builds roughly 7 million doors and 1.5 million exhaust systems a year. It is also a tooling company and provider of automotive tooling facilities.
The chosen challenge
Hirotec America, the Hirotec shop in Detroit, was chosen as the project test bed because of the significant integration challenge posed by the unique data types generated by eight different kinds of computer-numerical control (CNC) machines.
"On the one hand, we’ve collected data throughout our history, both on the production side and the tooling side. But although great emphasis was placed on collecting data, not only weren’t we getting all the data from tooling company clients, our own data wasn’t in a form needed to support timely business decisions," Hester said. "The quality assurance and maintenance departments placed great store by their data, but didn’t share it with each other in a meaningful way."
One result of having these silos of information could be, for example, that managers didn’t realize until too late that the defects found on some manufactured doors were caused by a concurrent robot maintenance problem. The real-time nature of the new IIoT solution would address this challenge.
In general, said Hester, the situation at Hirotec was much like that found in many plants today. Almost everyone is familiar with the "Tuesday morning meetings" that take place in many manufacturing or production facilities. In these meetings, everyone involved gathers around the prominently displayed key-performance indicator charts to review Monday’s production. Important things come out of those meetings, "But it’s a simple fact that by then it’s too late to address Monday’s challenges," said Hester.
PTC, a computer software and services company founded in 1985 and located outside Boston, provided the solution used by Hirotec. PTC is well-known for its computer-aided design (CAD) and product-lifecycle management (PLM) solutions. Its acquisition of ThingWorx in 2014 and Kepware in 2016, as well as that of Adexa and ColdLight, position PTC as a significant player in emerging IIoT technology markets.
In ThingWorx, PTC has an analytics platform developed by a team with deep roots in the production industries. Kepware brings solutions from a company with many years’ experience integrating an installed industrial technology base with the latest control and computing elements.
The IIoT PTC solution at Hirotec includes the following elements:
ThingWorx IoT Platform—set of integrated tools and capabilities for IoT solutions, including analytics for real-time anomaly detection at the edge, predictive analytics, and simulation.
Kepware KEPServerEX-solution for device-to-cloud interoperability, acting as a single source of industrial automation data to multiple applications allowing users to connect, manage, monitor, and control diverse automation devices and software applications through one intuitive user interface.
IoT Gateway for KEPServerEX—agent leverages the 150+ communications drivers within KEPServerEX to move real-time industrial data on the edge into ThingWorx.
The PTC solution gives Hirotec technicians, engineers, and managers access to real-time, role-based visualizations of the plant floor network’s current state, including electronic and industrial devices.
"Kepware is how we get the data into the ecosystem," said Hester. "ThingWorx is our engine for data analytics and machine learning. We like that ThingWorx can either be on-premise or in the Cloud." said Hester.
Hirotec already made use of Windchill, PTC’s PLM platform. It was therefore easy also to connect to engineering math data, said Hester. "At that point the idea of the model or a digital twin as a management concept becomes very real for us. PTC also has an augmented reality tool—Euphoria—that we can use to train operators and technicians."
As an implementation task, "IIoT isn’t like ERP," Hester said. "It matches the way we do business, not the opposite, and that’s very important."
Supported by PTC, Hirotec assembled its IIoT framework using the SCRUM method, aimed at achieving fast implementations via "agile" sprints. In this case, the sprints were of 6-week durations.
SCRUM methods are typically applied to software development as an iterative and incremental framework. A key principle is that project parameters will change during development, and thus progress can be more important than initial comprehensiveness. The method delivers substantial benefits, Hester said.
"We approached it as a software project, not a capital hardware project," Hester said. "We fund these smaller projects and use them as a proof of concept. In each instance, we may spend hundreds of thousands of dollars, but we’re not committed to spending millions."
An ancillary benefit is that to move forward, "we don’t necessarily have to elucidate a vision for where the company will be in 10 years. There’s flexibility this way, and it’s easier for interested parties to understand what we’re doing," Hester said.
As mentioned, of special interest to the initial project was to demonstrate that meaningful integration could be achieved across all the different CNC machine types at Hirotec America. Besides being of different types, the 60 machines were also of different vintages. Thus, although CNC technology has only been around for decades, given the current speed of technology development, the machines already represented different technology eras.
It was easy to connect the different machines’ device protocols into the common system, Hester said. "Kepware just does it, so there wasn’t any reason for us to dive further into it than that."
However, Hester noted, it’s a truism that "it is the machine that breaks that everyone knows best. We just put everything into Kepware. It worked so well that we didn’t bother to map out the network. We didn’t realize our mistake until we went to add a machine and didn’t have a port for it. Then we had to go back and create the needed network map."
One internal stakeholder in the project that Hester and his colleagues were intent on getting support from was the information-technology (IT) department.
"It was helpful to engage with IT up front. So often, too often, information technology departments have projects thrown over the wall to them and then they are told it was needed yesterday," Hester said.
By engaging with the IT department immediately, departmental associates were more supportive of the project. "Firewall and server issues were addressed expeditiously, servers came on line quickly, connection to the cloud was made in less than a day," Hester said. "Audits didn’t have to be performed after the fact."
Another set of stakeholders are the system users. As with the CNC machines, stakeholders are of different types. "If you’re investigating an occurrence, the analysis is timelier and more complete if everyone finds the data in a familiar context," Hester said. "For a design engineer, that means a CAD or PLM platform, for the operations manager that means SCADA or a manufacturing execution system. The ability to do that for the users was a benefit."
The interface provided both tracks machine uptimes and profiles the complex machine states involved in a way that can be intuitively grasped by operators, engineers, and managers. Engineers can compare the performances of different machines. Investigations can uncover linkages between maintenance anomalies and quality, operations, or production challenges.
Clearly, however, what Hester likes is that "we have a solution that can be scaled to address different challenges."
Hester is now working on a second IIoT project. At Hirotec Japan he is helping to develop an IIoT application that involves 50-point inspections of advanced robotic systems.
Industrial suppliers of many stripes, Hester noted, are introducing IIoT solutions aimed at specific maintenance or operations challenges. What Hirotec wanted was an eco-system platform that addressed specific challenges but that also made strategic sense.
With the solution, Hirotec America increased visibility into its CNC shop processes and has deeper insight into operations. "With small IIoT projects there will mainly be ‘soft’ benefits," Hester said, "although we’ve measured what we’ve accomplished. Once we were done, the plant and operations managers wanted it on their desktop immediately, as well as a display on the plant floor."
Having CNC machine uptime data improved shop scheduling, based on a solid understanding of past and current states. Manufacturing leverages real-time shop-floor data as an input to the ERP scheduling module, optimizing parts flow to the CNC modules.
Greater insight into asset and resource allocations based on smarter questions about priorities helps determine the most effective course of action. The result is improved productivity at Hirotec America.
Kevin Parker is a senior contributing editor at CFE Media.
– See other articles from the supplement below.
The newly implemented system melds operations and maintenance data, and by doing so, it can help employees identify trends that lead to contextualized, actionable insights.
By engaging with the IT department immediately, departmental associates were more supportive of the project.
The analysis is timelier and more complete if everyone finds the data in a familiar context.