There are many sides to maintenance in a manufacturing plant. Some involve mechanical components, and others define the procedures and processes needed to service them. The list of mechanical considerations is long, including material selection, surface treatments, bearing selection and implementation, pneumatic actuator sizing and selection, fixture geometry, serviceability and critical adjustment features.
The control system is just as important of a piece of the maintenance puzzle as the mechanical side. The control system itself can aid in machine maintenance, and there are tools available that help to ensure a long and productive machine life. And that is just to get started. The controls hardware and programming can help to maintain a machine in many other ways, as well.
"On the controls side of maintenance, the selection of quality, reputable components is almost always worth the extra expense," says Leon Krzmarzick, director of engineering at Delta Technology (deltatechinc.com). "Many manufacturers provide the tools needed to properly size and apply their components. Servo-motor suppliers have highly detailed software for properly sizing drives and motors. Robotic manufacturers have simulators that can accurately predict robot performance before any hardware is purchased. By using these tools, an engineer can design a reliable solution and ensure years of trouble-free service. Of course, success is only achieved if the engineer has a thorough understanding of the dynamics of the application and then includes some best practices in design and programming while keeping maintenance in mind."
Maintenance, not repair
As often happens, when a machine is not maintained, it can fail, so enforcing the maintenance through supplier vetting, component selection, design and procedures are important.
"In some cases where end users have neglected the preventive-maintenance schedule, it can have a negative impact on the bottom line and result in premature failure of some of the components," says Dean Colwell, controls manager—assembly, welding & AGV systems at Fori Automation in Shelby Township, Michigan. "The premature component failure often occurs when an end user moves the trained staff to other assignments and or neglects the preventive maintenance schedules."
Both the supplier and components can improve maintenance. "As a machine-tool builder, we have strategic component suppliers that have met our expectations for quality, longevity and overall performance," says Colwell. "In many cases, the end user may also specify components that they have positive experience with and that have met their reliability objective. Discussing maintenance and accessibility early and often during the design phase will ensure that it is kept on the forefront (Figure 1)."
Possible errors and end-of-travel sensors can help to speed troubleshooting and reduce damage to the equipment. "We do what we call a what-if analysis early on in the design phase," says Doug Putnam-Pite, director of software development at Owens Design in Fremont, California. "We break down the tool into sub-systems and write down the possible error situations that we think may occur. For each error situation we define the expected tool software response to the error and what we expect the user to do to recover from this error. This process helps to define sensors that may be missing from the design that are required for the software to detect and potentially recover from these errors. This process also brings to light error situations that are not recoverable with software with or without sensors."
It's important to not have maintenance activities turn in to repair activities. For the pneumatic actuators in the tool, Owens Design also makes sure that there are sensors available to show both the actuated and unactuated positions of the actuator for the software. For servo- and stepper-based actuators, it typically adds limit sensors to prevent crashes in automatic mode or operator error crashes in manual mode. These sensors help to prevent damage to pneumatic and motor-driven actuators when maintenance personnel are troubleshooting these components.
Some OEMs always try to design machinery with as little required maintenance as possible. "Lower parts count generally results in less maintenance," says Mike Krummey, electrical engineering manager at Matrix, a ProMach product brand located in Saukville, Wisconsin. "If you design the machine using as few components as possible, that design will be more reliable than one that has great complexity and a high parts count."
Control of maintenance
A few ways that a control system can improve machine uptime is though retries and keeping a count of cycles. "Retries are usually a symptom of inconsistent product, but nonetheless, there are times where a two- or three-time retry on a motion can allow the tool to keep running," says Krzmarzick at Delta Technology. "To get ahead of maintenance, counters for specific machine processes can trigger preventive maintenance warnings or alarms, enabling repairs to be scheduled and then made during planned downtime.”
A kinder and gentler control can reduce maintenance, as well. "In servo-motor applications, motion profiles can be softened if cycle time allows, reducing stress on the servo-axis components and the related assemblies," says Krzmarzick. "Additionally, the torque on servo applications can be limited, thereby reducing the chance of damage to the machine in an abnormal situation. In vision applications, techniques can be used to allow for fluctuation in ambient lighting or to handle degradation in the vision lighting."
The hardware itself can improve maintenance with an example being the use of shielded variable-frequency drive (VFD) cable versus a standard thermoplastic heat- and water-resistant nylon-coated (THWN) stranded wire. "THWN wire is certainly less expensive and more readily available," says David Paul, engineering design manager at Maverick Technologies a member of the Control System Integrators Association (CSIA). "However, in the long term, the 600-V insulation of THWN wire will fail when applied on a modern VFD due to the drive’s output voltage that exceeds 1,000 V. Shielded VFD cable has 1,000-V insulation, as well as shielding to minimize electrical noise generated from the VFD."
Installation of safety contactors for VFDs is another example of how hardware can reduce maintenance. "Earlier generations of VFDs required the installation of safety contactors in the line-side power to the drive," says Paul. "The latest generation of drives should have the safety contactors installed on the load side—between the VFD and the motor. One manufacturer’s documentation states installation of contactors in the drive line-side power will shorten the life of the drive. Designers must keep abreast of technology changes that impact best design practices."
Proper sizing and selection of motors, drives and power transmission components is critical, says John Kowal, director, business development at B&R Industrial Automation. "Use the manufacturers' servo cables," he says. "The terminations present many potential failure points if you are building or sourcing cables from third parties. Make sure the cables are properly rated if they are moving with a machine member, as they do on a gantry mechanism."
Simple best practice
Keep maintenance simple. "To minimize and prevent power surges and transients in control panels, install transient surge suppressors on all coils," says Paul at Maverick Technologies. "Installing a line filter on incoming power at the main disconnect in control panels is a simple way to reduce component failures due to electromagnetic interference, voltage transients, power surges and lighting strikes."
Machine components with quick and error-proof component replacement is another method to simplify maintenance. Paul gave an example of how a machine that uses proximity sensors could be equipped with sensors that have a quick disconnect, instead of a molded cable, making the sensor and cable two separate parts. "While the sensor and quick-disconnect cable add cost to the initial build, they save in the long run by reducing downtime and maintenance cost should the sensor need to be replaced," he said. "Error proofing the replacement of the sensor assures the machine will restart and secondary damage will not be inflicted with a miswired replacement part."
There are many ways technology can help to simplify maintenance, says Kowal at B&R Industrial Automation. "This includes safe motion in drives to aid in jogging machine components, augmented reality and virtual reality, on-screen work instructions using video and immersive 3D animation, and QR codes at maintenance locations that allow the technician to download those instructions to his/her mobile device," he says.
Predictive control and monitoring
"Preventive and predictive maintenance can be provided by control system programs, running either locally or remotely in the cloud," says Simone Gianotti, EcoStruxure industry business development manager at Schneider Electric. "In the past few years, there has been a lot of talk about preventive and predictive maintenance. Preventive maintenance requires either scheduled maintenance activities to happen before anything goes wrong or when some anomaly occurs such as a motor that starts using more current than usual."
Predictive maintenance, the direction that the market is slowly moving toward, requires a deep understanding of the machine behavior and sophisticated algorithms that run statistical analysis on the machine, says Gianotti. "The goal is to be able to identify trends within the machine and can predict when the next failure will occur and which components will be involved," he said. "Once applied, predictive maintenance results are more accurate and more effective than preventive maintenance."
The monitoring capabilities of a control system is another important part of improving asset reliability. "A control system can be set to monitor process conditions that would lead to equipment damage and adjusting the process to more equipment friendly conditions," says Trond Straume, chief technology officer at Schneider Electric AVEVA group. "A control system can also be set to monitor the equipment itself. You can monitor temperature, pressure, vibrations, noise—all that could indicate problematic equipment conditions."
Control system programs will require the right sensors and using the data that can already be extracted from machine tools, says Joseph Kernich, consulting application engineer, CRL at Siemens. "Smart sensing technology can also be utilized during traditional machine warmup cycles," he says. "Having the ability to provide intelligence about how a servo axis reacts during cyclic moves gives operators information that they can use to adjust lubrication and mechanical wear automatically. These same tests can be archived and trended over time to predict machine component wear."
Build a model
"Modeling of the machine design life and maintenance schedule based on the usage can reduce the maintenance and improve reliability," says Jason Tsai, vice president, product development, at Fanuc. "For example, a typical grease change is recommended at 20,000 running hours. However, the grease life can vary significantly based on the operating conditions."
Proper modeling of the grease life for a given machine under different operating conditions can be complex and critical to reduce the maintenance cost and improve the reliability, explains Tsai. "A control system can be designed to model the machine usage in real time and provide the proper maintenance schedule," he says.
A typical production environment has many machines from different vendors. "The maintenance schedule will vary depending on each machine design," says Tsai. "To properly manage the maintenance schedule, the control system needs to receive the production running data from all machines, perform the maintenance analytics based on the production running data and generate the maintenance report for each machine."
As there are typically more than 100 moving parts in a typical robotics application, the modeling of all moving parts to provide an accurate maintenance schedule can be very difficult and challenging, continues Tsai. "Maintenance requirements can also vary significantly depending on the operating and environmental conditions such as temperature, humidity, motion aggressiveness, high-speed emergency stop frequency and system loading," he says.
Artificial intelligence can offer a solution to address this complex problem, continues Tsai. "Industry 4.0 drives the smart factories by moving the production machines’ data and environment data to a server or a cloud," he says. "The analytics, which analyze the production running data and generates a maintenance schedule, can now be based on large production data from all machines in order to improve the algorithm constantly in real time. The analytics provided in a maintenance report can not only reduce the production maintenance cost, it can also feed back to the machine vendor to improve the mechanical design, reduce maintenance costs and improve product reliability."
Having the proper hardware and programming in place can go a long way toward simplifying machine maintenance, and there are many best practices, as well. The next step is collecting the production machine usage data. This will allow improved artificial intelligence analytics which will be significant for the future of the automation industry.