A key success factor for Industry 4.0 and IIoT (Industrial Internet of Things) initiatives is the emergence of more and better sensors in machining centers, and even in the cutting tools themselves. These sensors provide the data and connectivity that are the foundation for the “factory of the future.”
But, far from being futuristic, there are a range of “smart sensors” available today—collecting data and showing operators the health of their machines and the metalcutting process. The evolution is achieved through increasingly accurate measurement of the position of the part and the geometrical form of the finished part, as well as the configuration and control of the tools used in the process.
Manufacturing Engineering asked a number of industry experts to assess the current state of sensor-based cutting control systems, and also to gaze into the crystal ball to help identify future opportunities.
In the area of sensor-based control and optimization, where are we on the continuum from drawing board to mature products? The answer to that question is, “It depends.”
“For basic measurement and adaptive control, we are well along the learning curve,” said Frank Powell, product manager for grinding products at Marposs Corp. (Auburn Hills, MI). “Machine tools can be equipped with a variety of in-process sensors and transducers. At the first level, these sensors are utilized for machine protection since a system can react 1000 times faster than an operator to an unexpected force strain or potential collision. The technology is advancing through increasingly accurate measurement of the position of the part and the geometrical form of the finished part, as well as the configuration and control of the tools used in the process.”
In terms of Industry 4.0, cutting tool digital manufacturing technology is far closer to the drawing board than a mature product, according to Neil Munson, technical sales specialist for Silent Tools at Sandvik Coromant (Fair Lawn, NJ). “We are at the precipice of a paradigm shift in our industry. In fact we are taking the initial leap into sensor-based, intelligent cutting tools at our company,” he said.
Joe Volansky of BadAxe Tooling Solutions had a little different view. “Products are proven and well along the learning curve, and can be applied in revolutionary ways,” Volansky said. “In our case, we are introducing a product for milling operations based on proven science, which presents a solution in a completely new way at the edge of machine physics.” BadAxe Tooling Solutions is an alliance of three US companies: Briney Tooling Systems (Bad Axe, MI), Fullerton Tool Co. (Saginaw, MI) and BlueSwarf LLC (State College, PA).
“My opinion is that we are in the middle of the curve in terms of development,” said Wade Anderson, product specialist manager and Tech Centers manager for Okuma America Corp. (Charlotte, NC). “For basic equipment efficiency and the ability to communicate with peripheral devices, there are a number of off-the-shelf solutions ready to go right now. However, artificial intelligence (AI) and augmented reality (AR) technologies are more up and coming.”
With the vast amount of data now being generated from equipment on the shop floor, the up-and-coming technologies that Anderson refers to are essential, according to Andy Henderson, vice president of engineering for Praemo (Kitchener, ON). “In the automation world, there has been a very ‘physical’ evolution since the 1970s,” Henderson said. “The next frontier is the ‘mental’ evolution—adding intelligence to the data using AI and machine learning (ML). On the continuum, these technologies are mature and are used continuously in the personal and consumer spaces (Google, Amazon, Facebook, etc.). Applications inside the manufacturing industry have lagged, but software using these technologies, like Razor, are helping machine shops realize the value of AI and ML.” Razor is a Praemo product that connects to data sources, gathers information, and provides insights and suggested recommendations to help users take preemptive action to reduce equipment downtime, optimize production processes, and manage risks.
Our industry experts all agreed on the key role that sensors play. There’s an old adage, “What gets measured gets managed.” Though it is uncertain who said it (some suggest Lord Kelvin, others Peter Drucker), it underscores the fact that any attempt to control or optimize a machine must be based on fast, accurate, and reliable data at the key point of contact—where the cutting tool meets the workpiece.
Technically speaking, technology to successfully collect, distribute, and analyze data for adaptive control was available long before the current interest in Industry 4.0 and IIoT. Historically, in-process measuring equipment can trace its roots back over 60 years (for example, Mario Possati’s gauge for checking parts during grinding). However, the major advances in sensors have occurred much more recently.
One of these advances is a boring tool that automatically compensates for cutting edge wear. “Regardless of how sophisticated a CNC machine tool is, it cannot automatically compensate for cutting edge wear on a boring tool,” said Anthony Bassett, president of Rigibore Inc. (Mukwonago, WI). “Our battery-powered ActiveEdge tool automates the in-process sizing of boring tool cutting edges. It uses wireless technology to remotely adjust multiple cutting edge diameters on a single tool, optimizing process performance and eliminating the need for operator intervention.”
Forging into new territory in the milling area is the BadAxe Smart Tool assembly and software system. BadAxe replaces trial-and-error with physics to immediately hit the “sweet spot” of a machine when it is started up, according to the company. Using measurements of the resident vibration frequency patterns for a specific combination of machine, spindle, toolholder, and tool, the company’s patented algorithms and analytics calculate the optimum stability zone for the machine to operate in. The result is an off-the-shelf, pre-balanced, and assembled toolholder and tool, plus a software interface optimized for individual machines.
“Everyone knows that vibration and chatter are major limiting factors in high-speed machining,” Volansky of BadAxe said. “Typically, operators use trial-and-error to find the right feed, speed, cut width, and cut depth. In many cases, they think they’ve hit the wall in terms of performance when, in reality, chatter-free operation would be found in a counter-intuitive setup where much higher metal removal rates exist. That setup is defined by our product immediately from start-up, regardless of step-over, full slot cutting, or corners. The underlying technology has been proven to significantly increase a job’s ROI by organizations such as Boeing Phantom Works, the US Department of Defense, and others.”
BadAxe is building a large database for different machine and assembly combinations. If a customer’s machine is not in the database, a one-time test can be done to determine its setup dynamics.Making the move from in-process to in-tool sensors is Sandvik Coromant with the introduction of its SilentTools+ intelligent cutting tools. “SilentTools+ are damped tooling solutions to provide tremendous insight into long-overhang machining processes where operators typically are flying blind,” said Munson.
“Long-overhang machining in a high-tech facility is challenging because [operators] want to have control of the process all the way to the cutting tip,” he said. “We have embedded sensors inside our SilentTools that communicate via Bluetooth with a graphics interface so that an operator has eyes and ear inside the enclosed machine and deep inside a bore. In addition, it has a feature that enables quick and easy setup by determining when the cutting tip is on center.”
Using the SilentTools+ system, tool deflection, chatter, cutting forces, tool load and temperature are monitored in real time. “Because of the insight into all of these areas, we think that scrap rates will be reduced and productivity gains will be realized, yielding real savings,” Munson added.
While some sensor output is for information only, the value-added is when sensor signals are processed and used to control part quality and tool condition. Manufacturers can minimize toolpaths and machining time, improve surface finish, maximize machine life, and efficiently machine more challenging parts, such as those with complex geometries, thin walls, hollow cylinders, and slender shafts.
Said Anderson, “The Okuma Monitoring System is a good first step. It allows up to 64 different inputs that can be customized to process-specific needs. We have customers who have been able to measure machine downtime and analyze OEE [overall equipment effectiveness] data to determine that a simple thing was shutting the machine down when an operator was tied up elsewhere. By using the monitoring system, the operator [can be alerted] that there is a potential problem coming that preventative action can avoid.”
Henderson of Praemo believes that the road to manufacturing success ultimately runs through the mountain of data that tools are generating in metalcutting applications. “These data are valuable ‘raw material’ for optimization,” he said. “The algorithms for analyzing data in our Razor software can sift through vast amounts of data to find interrelationships and patterns that a human with a spreadsheet would never uncover. AI and ML technologies are used to develop models that help operators and managers arrive at clear conclusions very quickly—and take action based on data-driven conclusions. While the intelligence is artificial, it very closely mimics how humans in the manufacturing world approach things. The only difference is that it doesn’t get overwhelmed by massive amounts of data. The longer data are collected, and the more data collected, the more Razor learns.”
Today’s systems have proven their ability to maintain the cutting process, anticipating when something is not right and notifying an operator or the machine to take corrective action, according to Powell of Marposs. “The decision to notify only, or take automatic actions, is dictated by the customer,” he said. “Technically, we have a lot of capabilities for automatic control that in some cases are not being utilized.”
A new system from Marposs called BLÚ is the result of over 50 years’ experience on grinding machines and other machine tools. It combines all the machine tool monitoring and process control applications in a single system for real-time tool and process monitoring, adaptive control, and crash mitigation with a fast processing and sampling time. Said Powell, “BLÚ is designed to integrate almost seamlessly with the machine tool to aid the operator in any mode: setup, manual, or automatic.”
The heart of the system is a master node housed inside the machine cabinet. It has a bus structure to plug in other nodes for on-machine measurements, acoustic emission monitoring, wheel balancing (grinding), tool retraction, and other functions.
The Zenith system is a closed-loop, automatic system for boring operations, according to Bassett of Rigibore. “Traditional systems are purely mechanical without continuous feedback to the boring machine,” he said. “Zenith provides feedback to the machine control to confirm that the tool has adjusted by the desired amount. This position sensor feedback prevents a bad part from being machined.”
“I don’t necessarily see an end game for digital or sensor-based cutting tool solutions,” said Munson of Sandvik Coromant. “In modern machine shops and manufacturing facilities, sensors and the use of IIoT in cutting tools will become more ubiquitous over the next five to 10 years, following the path of consumer [product] applications. For us, the natural progression is for integration of our SilentTools+ into the controls of CNC machinery. Today, it is a stand-alone system with a live process view. We are working with machine tool builder and CAM software partners to move from a stand-alone system to machine control and process integration.”
Anderson of Okuma agreed. “Technology and, most importantly, the implementation of technology, will breed new technologies,” he said. “I thought many years ago that an auto-dimming rearview mirror in a vehicle was ridiculous. Now, I wouldn’t buy a vehicle without one. In a similar fashion, in our industry new technologies will help develop new markets for alternative advancements.”
Powell of Marposs concurred that there should never be an end game. “I think that the technology is in place to monitor all the key parameters and variables,” he said. “Certainly, there are areas that can be improved, and we are working daily to develop sensors that are faster and more sensitive, and software that can process the sensor data more quickly and intelligently. Within 10 years, I’m sure there will be at least one game-changer, a technical breakthrough that makes a machine shop more efficient, more productive. When the game changes, we’ll change with it.”Volansky of BadAxe thinks that his company has a game-changer right now. “The pace of development is increasing,” he said. “Customers don’t want to wait five or 10 years. We have a product that redefines efficiency and helps machine tools operate at the limits of physics. The next step for us, though, will be to remove our tooling assembly from the equation by embedding our proprietary vibration sensing and data analysis right into the machine and its logic.” This way, the machine will only operate at its most highly optimized state, without operator trial-and-error, he added.
As Bassett of Rigibore said, “Some development targets for us are automatic ordering of carbide tooling using the CNC control, a shift to less intervention from operators, adjustments made from one central location, and the input of a solid model into the CNC machine while another process is in operation. The major limits we face today are the incompatibility and lack of standards between different CNC machine builders and machine controller manufacturers.” In the end, a proprietary approach will limit growth and is ultimately more expensive for a customer who wants to easily add new features, Bassett concluded.
According to Henderson of Praemo, some people define lights-out manufacturing, with no human intervention, as the end game. “Personally, I think of lights-out as similar to the quest for world peace,” he said. “For the manufacturing industry, it’s a concept that is attractive to strive for, but will not be attainable in many situations. Sensors built into tools and tool networks; software that seamlessly collects, organizes and analyzes data; and machine tools that are networked into a source of factory intelligence bring us much closer to that lights-out factory. This doesn’t mean that machines replace people. It frees them to do work that is less menial, more satisfying, and definitely more productive. It’s similar to the advent of the CNC, which automated the manual control of machining centers. It allowed people to work safer and more productively.”