How AGR Derisks Automation: Phased Development for Success

For instance, OEM and system integrator Arthur G. Russell Co. (AGR) in Bristol, Conn., always had an internal R&D and mock-up area, which was part of its due diligence process when clients requested a specific job. AGR builds and integrates assembly and packaging equipment from pilot to large-scale, high-throughput (20-1,800 parts per minute) for pharmaceutical, medical device, and other industry clients. However, its machines typically contain multiple flow, temperature, pressure, and other small process applications, such as measuring reagents down to microliter levels. However, more customers are also seeking to take automation in derisked stages, starting with proof-of-concept (PoC) and pilot stages.

Test small before scaling up

Because many clients need to perform inspections as part of their equipment development and production, Romano reports that AGR’s PoCs often used to include mocking up vision inspections, selecting lenses, lights and cameras in different applications, and writing evaluation reports, so they could eventually be rolled into full systems. Now, customers are asking for prototypes complete with accessories like fixtures for holding and assembling products, so they can quickly translate them into pilots and full-scale production.

“We very recently worked on a medical device assembly system, and the client wanted a tabletop PoC for a process that provided the full method for how one station was supposed to work,” explains Romano. “It was challenging to manage its parts, but the feasibility wasn’t completely known since no prototype was completed previously. We had to demonstrate the operation of this challenging station. This PoC was good practice for the customer and AGR.”

Romano adds that clients are also seeking to derisk projects due to multiple, former failures of large assembly machines that caused low production yields and poor end-product quality. “Many companies have reported to us they’ve been burned by going straight to full production, where costly machines had to be scrapped or substantially reworked when it turned out they could only run at low overall equipment effectiveness (OEE).”

At the same time, clients want to derisk because shorter product lifecycles and increasing demand for variety are squeezing available product development and commissioning times. Along with other constraints imposed by retooling to make newly developed products, these efforts also severely compress related schedules. This requires OEMs, system integrators and their clients to be flexible enough to devise and adopt other innovative strategies, according to Romano.

To provide concepts and products that achieve goals like improved design for service, adaptability, maintenance, continuous improvement and data gathering, Romano reports that more components are being designed with automation in mind. These designs strive to:

• Reduce part counts;
• Minimize handling and reorientation of assemblies as they’re made;
• Feed and join parts more easily and consistently:
• Inspect and test parts more easily and consistently; and
• Determine if these tasks can be automated.

“Standardization can be a big help, but every project is different and still needs custom automation to satisfy specific assembly needs,” adds Romano.

Romano reports that derisking any process or product must be done progressively, so each revision at each stage can be thoroughly evaluated and further tweaked if needed. “Derisking must be done in steps, so we can make certain to prove each one,” adds Romano. “We’ve done several medical device projects in the last couple of years, and doing the PoC was essential in almost every case. A derisking project can take up to three to 12 months, depending on size and complexity.”

Just as standardized and modular components simplify and streamline its projects—and reduce unexpected variability and risk—AGR employs a standard method for PLC, HMI, and robotic programming and communications. Part of this is using common structures for programming, PLC tag names, and setting up PLC communication tasks. This methodology lets users transfer and share data between machines in a more straightforward manner, makes it easier to anticipate what results can be expected on the other side of a process, and lets users coordinate handoffs between Rockwell PLCs and Epson robots that AGR partners with via EtherNet/IP protocol and networking.

AGR also employs a standardized, user-defined tags (UDT) structure, and a library of add-on instructions (AOI) and subroutines. This lets them approach object-oriented coding methods with a library of functions that can be easily dragged down and configured to bring programs online with less programming and faster debugging. This ease of use helps with all project stages and derisking from PoC to full-scale automation.

Beyond today’s immediate projects, Romano reports that derisking efforts and the standard and modular capabilities they encourage can likewise be helpful during a component or process’ entire lifecycle and enable a client’s complete automation roadmap.

“Developing a new product or process from scratch obviously takes a lot of R&D, so it’s useful if we can use a solution that’s already been proven, and determine if it can be scaled up,” explains Romano. “As for roadmaps, where the customer is at with their process or business dictates much of the type of machine or capability they’ll need. There are entanglements here because, if the company doesn’t have a full maintenance staff, sufficient spare parts or available automation support, then trying to implement a sophisticated machine will likely be less successful. This is why workforce shortage and skills-gap issues must also be addressed. Machine operators, maintenance technicians, and process and manufacturing engineers must be ready to provide maintenance and support, and handle machine faults and related troubleshooting.”

Romano adds that AGR can also expand its outreach to do maintenance and troubleshooting, but it’s more efficient and less costly if users can close their skills gaps and have inhouse maintenance and support. “For example, if a client has a semi-automated or fully automated machine on their plant floor, we ask how they’re currently handling maintenance and support,” says Romano. “Do they need us to provide maintenance services as part of their ongoing automation process?”

While evaluating and derisking, other questions arise, such as can a client go 100% automated, or do they need manual steps between some work cells? “There’s also the question of cost because a completely automated machine may cost several million dollars, while the semi-automated machine may be a fraction of that,” adds Romano. “The question is, what’s their need? Once we have an answer, we can use this knowledge to provide development and derisking that more closely matches their budget and inhouse support capabilities.”

The three most common approaches to project selection are:

• Payback period determines how long it would take to produce enough revenue to recover the initial investment. Methods include percentage return on equity (ROE) invested, percentage return on invested (ROI) capital, and return on asset (RoA) or how much profit it can generate. Romano reports the typical payback period varies from six months to two years.
• Internal rate of return (IRR) is the expected return on a project. If the rate is higher than the cost of capital, it’s a good project.
• Net present value (NPV) is the sum of all future cash flows over the investment’s lifetime, discounted to the present value, which may be the most effective method.

Because answering these and related questions requires research, Romano adds that it helps to have a well-aligned technology partner do some of the R&D and specifications.

What makes a good product development and derisking partner?

Based on its own experiences and input from its customers, system integrator Arthur G. Russell reports the most valuable traits of an aligned technology partner can be identified by asking the following questions:

• Do they have a lab for developing key processes?
• Are they capable of starting with manual and semi-automated processes, and is this capability affordable?
• Do they have the ability to do a pilot scale?
• Can they scale to full production?
• Do they have in-house capabilities for all design and production needs?
• Do they offer the proper solution for the job? (For example, are multiple platforms available?)
• Where are their capabilities and supply chains located? (Footprint considerations are important for reshoring.)
• Do they balance robust, proven technology with new technology, and apply the proper one?
• Can they match the proper level of technology that the customer can handle?
• Do they create designed with customer capabilities in mind?
• Can they help develop user-required specifications (URS)?