Factory-worker-checklist-hero

What's the best practice for factory acceptance testing?

May 28, 2020
Five industry experts weigh in on how to decide what to include in a FAT

A Control Design reader writes: As an automation engineer at a large discrete part manufacturer, I need to create a factory-acceptance-test (FAT) plan for a $2 million high-speed, automated system. It has multiple stations with several robots, motion control, vision inspections and many fieldbus digital and analog I/O. The question is what should I include in a FAT? I don't know where to start. Also, are there diagnostic tools or software to make performing the defined tests easier to complete? What do you suggest?

[javascriptSnippet]

ANSWERS

Built-in diagnostics

Many automation technology manufacturers include testing and validation processes in their integrated development environments. The embedded processes guide users through a step-by-step process to validate different operations and components to ensure that they work as expected.

TODD MASON-DARNELL / marketing manager, services & safety / Omron Automation Americas

FAT the specs

You will want to attempt as much testing as physically and economically possible at the location of testing. The main goal of the FAT is to certify that what has been built is able to perform the functions that are possible to test and is the milestone that will determine whether to ship the system to site. If the FAT is thorough and no changes occur other than installation, you can isolate the factory-acceptance-tested (FATed) system as an area that is highly unlikely to cause an issue during commission. In essence, the problem should be how the FATed system was electrically/mechanically connected at the plant site. The FATed system is not always going to be a part to avoid blaming issues during commission but should be one of the last areas to check, unless the conditions presented lead back to it.

To start, begin with the various design documents. Functional specification and other design specs along with mechanical and electrical drawings are what should determine the tests to be performed. For example, electrical drawings for a panel should be checked for each wire to be correct from the hardware in the panel to the terminals where the field wires will be terminated. This is often done by the shop that builds the panel, and, depending on who the system is developed for, one may need to witness this test or redo the testing in front of the customer. The specifications will detail how hardware should be utilized and programmed. This will need to be tested to assure the specifications have been followed. This portion of the test is where the physical and economic judgment of feasibility comes in. If you are staging this in your office, you will likely be short on space, power and quite a few other things. Here you assess what can be tested. Some of this testing could be handled with simulation. Simulation can be done with software or using potentiometers, switches, dials and meters to simulate or read signals. If you system-assemble items and you have enough space to only partially test, and it is physically possible, you will want to demonstrate that in the FAT.

So, what do you gain from completing a thorough FAT?

  1. Prior to going to the field, you likely catch a high majority of issues in all areas of the design and may have some unforeseen issues as per what the end user wanted remedied.
  2. Commissioning should be considerably shorter if a thorough FAT wasn’t completed, as most problems are arise during field installation.
  3. Most importantly, changes and corrections that occur in the field are considerably more expensive, so the savings in money and field time will lead to a happier end user and field team. 

DEAN DEGRAZIA / operations manager / Applied Control Engineering (ACE) / Society for Maintenance and Reliability Professionals (SMRP) member

Envision all scenarios

Most of the time the FAT is performed to see if a machine meets the original specification. The specification must be well-written, and the machine must be thoroughly designed for all production scenarios. Any design changes required after the FAT can create extra costs and have the potential to ruin the project budget.

To make a good factory acceptance test, you must consider all possible machine functions. I would avoid using extra software to evaluate the machine for the FAT. Instead, design a FAT that encompasses all possible production scenarios and use the software already available in the HMI or PLC.

I would start by breaking down all machine functions separately and verifying that all parts of the process are working correctly. Then I would run product. However, just producing product is not good enough for a FAT because the machine must be able to handle all raw materials given to it, even if the material is out of specification. The FAT should test samples of raw material that meet the design specifications and samples that are outside of the tolerance range. A properly designed machine will reject irregular materials and accept those that meet the specification. Vision systems, smart sensors and mechanical test fixtures support this.

Redundancy must also be considered. Is there a way to shut down one part of the machine while the rest keeps running? Is there a way to substitute a human into the process while a part of the machine is being repaired? In short, all conceivable scenarios must be taken into account to verify that the machine will meet the requirements of the factory.

TIM CICERCHI / technical services manager, factory automation division / Pepperl+Fuchs

Step by step

The focus on the process of running a FAT is as important as testing to specifications. Some key elements in order of occurrence in a FAT process are as follows.

Clear operational specifications to be defined up front and agreed upon:

  • expectations to vendor to be clarified before FAT
  • confirm standardization of features and specifications to other systems on-site.

Break system out to major sections:

  • test each system individually
  • confirm data availability, communications and access from each part of the system.

Functional operation to be validated:

  • clear test plan that is communicated to vendor up front
  • acceptance criteria for the testing to be performed
  • define needs of the actual FAT execution, such as tools, labor, test products to use
  • testing performed to be representative of actual production usage, using actual products and throughput requirements.

Utilize a sign-off checklist:

  • complete as-built electrical and mechanical drawings
  • proof of certification, such as UL and NEMA,
  • user manuals
  • controller programs, if included contractually
  • operations and maintenance guides.

The usage of diagnostics tools would be based on specific requirements of the system. Some of these tools may already be in the production environment. Some of the tools may be provided by the vendor. This would have to be defined when the validation plan is defined prior to FAT, to set the right expectations from the vendor.

CHRIS THOMAS / senior application design engineer / Schneider Electric

Use best practices

Best practices for FAT should typically be categorized into multiple sections such as:

  • hardware
  • software
  • HMI/SCADA
  • field device communication test 
  • cybersecurity.

Each of these parts should then be divided in visual and functional tests. It should be reviewed and approved by stakeholders involved in the test. The FAT plan should clearly detail test description, which can be broken down into multiple tests, and each test should have acceptance criteria so there is no ambiguity on what qualifies for a successful test.

In addition to testing different areas such as hardware or software, it is essential to perform integrated FAT with all components connected. If the test involves integration with drives or servos or other critical components, it is recommended to have at least one sample of each type available for FAT to ensure that communication with these critical components will function as required when installed on a machine.

Most suppliers of control systems or PLCs also offer simulator, which is a good tool for application software and HMI tests without hardware. Some automation vendors offer simulator on a cloud platform, which can also enable remote software FAT with a customer, vendor or other parties involved helping to reduce travel needs.

SWAPNIL V. ADKAR / senior market development manager, global OEMs / Honeywell Process Solutions

About the author: Mike Bacidore
About the Author

Mike Bacidore | Editor in Chief

Mike Bacidore is chief editor of Control Design and has been an integral part of the Endeavor Business Media editorial team since 2007. Previously, he was editorial director at Hughes Communications and a portfolio manager of the human resources and labor law areas at Wolters Kluwer. Bacidore holds a BA from the University of Illinois and an MBA from Lake Forest Graduate School of Management. He is an award-winning columnist, earning multiple regional and national awards from the American Society of Business Publication Editors. He may be reached at [email protected]