Many manufacturers of machine tools such as lathes and machining centres offer multi-machine networks and management functions, but this is of little use for those running a mixed fleet. So what options exist for engineers keen on getting a closer handle on machine tool performance, across the board?
Among the new entrants to this market is FourJaw, a Sheffield-based tech spin-out with cloud-based data capture and analysis technology that claims to unlock five-fold increases in productivity. Co-founder and CEO Chris Iveson explains its story.
“My co-founder and CTO Robin Hartley created a clever piece of software to help research engineers at the University of Sheffield’s Advanced Manufacturing Research Centre [AMRC] monitor the many different machine tools on site, which include DMG Mori, WFL, Starrag and Makino brands. It helped researchers optimise speeds and feeds, and monitor tool loads. Such was its success that we decided to spin out in September 2020.”
Following a period of fundraising, which now includes a significant six-figure investment from a combined angel and venture capital syndicate that values the company at £2 million, the FourJaw solution is ready for commercial sale.
“We’ve done our research and created a product that is vendor-agnostic, quick to install and inexpensive,” says Hartley. “Talking to industry, we discovered that few were interested in knowing the exact speed of their spindle, for example, but instead wanted to know about machine utilisation, cycle times, the reasons for machine stoppages and corrective measures, and how to improve.”
CHANGE OF APPROACH
Originally, the FourJaw solution adopted a sensor-based approach, but this proved painstaking in terms of installation time, so the pair developed a straightforward plug-and-play device that is despatched in the post for self-installation on any machine, regardless of manufacturer. Based on a software-as-a-service (Saas) pricing model, the upfront costs and risks are minimal.
The low-cost hardware device, essentially a micro-computer, sits on the machine’s exterior and connects via a cable to the electrical cabinet. Inside the cabinet, a proprietary circuit board hosts a pair of current clamps, one of which hooks around the machine tool’s main power supply, while the other goes to the spindle motor. Users simply enter the Wi-Fi password on the device to start monitoring. A restricted version of the software is available free-of-charge for up to three machines so that users can evaluate the product before committing to the fully-featured software.
“Most CNC machine tools register utilisation figures lower than 30%; FourJaw provides a way of finding out why,” says Hartley. “Just a 10% increase in productivity would be revolutionary; but our device delves deep into the brains of shop-floor machines, decoding data to drive productivity gains well in excess of this.”
IN MINT CONDITION
When it comes to the monitoring of machine tool operating conditions, involving parameters such as vibration, temperature, sound, magnetic field and current, vendor-agnostic solutions are again available. For instance, German lathe manufacturer Weiler now offers its machines equipped with a condition monitoring system from IFM Electronic. The difference is that IFM supported Weiler in developing and implementing a system that is in fact suitable for all machine tools, regardless of type and manufacturer. The solution detects, monitors and analyses the process, machine and manufacturing data of any number of machines at different locations, with all information displayed in browser-based applications, making it accessible via any PC, tablet or smartphone.
“We’re finding that with the gradual disappearance of time-served engineering skills, customers want more integration with the data,” explains Vince Burson, condition monitoring sales & support specialist at IFM. “Most people don’t want to look at complex data, but instead prefer a visual format and email notifications. We have interface cards that communicate directly with the machine control/HMI and show users exactly how the machine is performing. There is no longer any need to be an expert in vibration to run effective condition monitoring technology.”
IFM can configure its devices to look for specific frequencies in line with the application. The data is fed to IFM’s new Moneo IIoT software platform where it is presented to users in graphical format, with alerts sent if pre-determined vibration limit parameters are exceeded. The system can also be adapted for other conditions, including current, speed and temperature.
“We can now use our new IO-Link VVB vibration sensors to feed an IO-Link master, before collating the data from these sensors to provide a holistic view of the machine, rather than just one process,” says Burson. “The days of having to scrutinise squiggly lines are long gone: our system will provide a graphical vRMS, aRMS, a-Peak, crest-factor and temperature – all from one sensor.” He was referring to a number of electrical engineering parameters, respectively: the root mean square of velocity; the root mean square of amplitude; the maximum excursion of the waveform from normal; the ratio of peak values to the effective value; and how hot it is.
Using a solution such as the VSA004 sensor on a machine tool, it is even possible to inform the user about specific issues, such as a loose or unbalanced bearing, for instance.
“We have experience of helping manufacturers that need help with a mixed fleet of machines,” says Burson. “Obviously, we consider each machine and build the solution accordingly, but with machine tools, we find most companies are looking for the same or similar information. With this in mind, we tend to use the same base configuration, introduce another module and simply change the frequency levels.” Two days is available with the first module bought.
BOX: GET THE EDGE ON MACHINING
The monitoring of cutting tool conditions has the potential to provide a plethora of useful information to machine shops. For this reason, the EPSRC is funding a £1 million research project headed up by principal investigator Zi-Qiang Lang, professor of complex systems analysis and design at the University of Sheffield’s Department of Automatic Control and Systems Engineering.
“Some researchers have estimated that the amount of machine tool downtime due to cutter wear or breakage is around 6.8%, while others put the figure closer to 20%,” says Lang. “Manufacturing costs can therefore be significantly higher than necessary, which is why the real-time and automatic inspection of cutting tool status and machine tool health is needed to profoundly address these problems.”
Lang is proposing a new paradigm for industrial process monitoring. The principle is to monitor changes in machining using the unique and physically meaningful frequency properties of a data-driven process model.
“This could fundamentally address many challenges, including the high noise/signal ratio problems associated with direct data analysis,” he says.
Regarding the automatic monitoring of the spindle tool set to predict cutter failure before it occurs, Lang’s team is working on vibration data. Research engineers are investigating how to reveal tool wear damage/defects by analysing the dynamic relationship of the vibration data measured at different locations. Where the features revealed by the data-driven model analysis are complex, the potential will exist to feed information into machine learning algorithms, thus streamlining the process.
“In these situations, a machine learning model may be trained to relate the process features extracted from the model frequency analysis to cutter status and machine tool working conditions, subsequently facilitating anomaly detection,” explains Lang.
Demonstration events are planned in 2023, the end of project funding, to disseminate the research outcomes.