by Julian Renz
Understanding the options to make the right choice for your operation
With all the talk of Industry 4.0, Internet of Things (IoT), and digitized manufacturing, machine shops might find themselves overwhelmed in deciding which options will work for them. A good way to find some guidance on these topics is to look at different products available and consider under which circumstances there is a benefit.
To start with, it is clear that part design, CAM and FEA have been digitized for a long time, and digitized manufacturing addresses the manufacturing process rather than technology tools.
Data transfer between devices refers to the manufacturing process, which is sometimes labeled “communication” of machines. Therein, communication ranges from simple I/O commands between machines, e.g., to load a part or infeed more stock into a lathe, to advanced applications that allow automated changes of NC code, based on data transmission by a PC. Profinet or Profibus connect industrial equipment in (almost) real-time using TCP / UDP to transmit data and Ethernet as data network. However, new solutions for IoT use higher communication standards to gather and share much more data. If data is to be distributed around a plant, a central server is required, which often runs on a Windows operating system. Indeed, having a common communication language is required for connected manufacturing, but the translation between machine languages continues to be a challenge.
With MTConnect and OPC UA, there are two established communication standards in North America and Europe respectively, and a lot of manufacturing companies and OEMs have bought into these platforms.
MTConnect tranfers data in HTTP format and outputs it in XML but typically needs an “agent” to extract data from machine tools using queries programmed in higher programming languages such as Python, Java or C#.
OPC UA is a similar standard, but it is more complex as interfacing with various other standards such as ISA 95 (interface between control and enterprise systems) or FDI (Field Device Integration) is possible. In addition to available adapters between these languages, an adapter between these standards and the OEM standard is needed. “Agents” and adapters as well as external hardware devices/switches on which these are embedded, have to be purchased. Heidenhain’s StateMonitor has MTConnect/OPC compatibility included, so that it connects to third-party controls.
In order to have total connection, the communication standards can be complemented by sensors. For example, a sensor on a milling machine can be used to monitor the pressure on a hydraulic fixture. Data from machine tools and data from sensors can then be fed into overarching production monitoring software offered by third parties. These software types are quite powerful, but not inexpensive. If all adapters and support for installation is provided, manufacturers still have to deal with a costly annual maintenance fee.
The question is whether the whole plant needs to be connected or just specific machinery. E.g. a tool presetter may only be interfaced with CNC machines to automatically transfer tool data which avoids manual errors and saves time.
Other viable solutions come from machine builders, and examples are a free connectors for a builder’s brand machines or free apps (a “barcode app”, a “camera inspection app’), developed together with software groups.
If we take a look beyond connecting devices in a plant, cloud servers are the next industry offering that are available. Cloud servers are useful for providing various plants access to technology and process data. Data clouds further add to repeatability of production systems, since proven-out processes can be stored in a common format and shared by planning, manufacturing and purchasing. Repeatability also refers to the established infrastructure itself, because once the cloud with all its data interfacing is up and running, companies will be reluctant to make changes since there was a high cost involved in the first place.
Cloud services are currently being offered by IT companies, CAD/CAM companies, machine builders and consortiums of OEMs. The latter is most interesting, as imagine your company procures machine tool, robot, and other devices from vendors of the same IoT platform, then the problem of having different data standards (languages) is solved and the connected system is less error-prone.
Any third-party equipment, however, still has to be added with software/hardware adapters that come with the issues that were mentioned above. Other questions are in regard to longevity of the common platform and more importantly, with respect to data security.
Who guarantees that data I’m sending to the cloud will not land in an outside entity’s hands (or servers)? How robust are firewalls of data clouds? Production data often is proprietary and the cornerstone of manufacturer’s competitiveness. Cloud services might offer to only provide data storage, but there isn’t a way to tell if someone takes a glimpse at the data.
Some OEMs take a proactive approach to this as they openly say they are using their customer’s data to generate technology databases, which they then provide to their whole customer base. This information might be useful as general advice, or to help small manufacturing startups on their feet with technology fundamentals. For every advanced machine shop, however, there should be some restraint in exposing its successful processes.
For manufacturers that are skeptical about cloud applications, a very functional way to achieve data-driven improvements in manufacturing are current state PLM solutions (e.g. DXF converters, order management), providing operator and engineers relevant documents at the right time.
Other OEMs offer improved after-sales service by collecting geographical data of customers. In that way, they are hoping to optimize service calls by locating nearby customers with equipment needs before service technicians are sent to remote locations for a single customer only.
Furthermore, sensors play a role in maintaining machinery and OEMs already have “predictive maintenance” solutions available. Think ball bearings for example, and the sounds they make if there is wear and tear or bearings are not perfectly aligned. Thus, predictive maintenance helps the OEM to optimize stocking of spare parts and prevents machine-downs for the end user.
Remotely, OEMs can monitor spindle power, send warning messages to customers and do utilization analysis in real-time, services for which they charge a fee. These monitoring offerings seem like a win-win for both vendor and customer, but there needs to be clarity about who sees and uses the data. Will competitors with the same vendor know about my shop’s efficiency? Who monitors my plant if licenses run out?
Alternatives to remote diagnostics allow disconnection after a service event or offer self-monitoring tools. Examples are spindle load functions or Heidenhain’s oscilloscope that monitors current, speed and acceleration control loops.
In any case, machine shops need to conduct their own due diligence to decide which connected machining products will work for them and which will not. As with conventional manufacturing equipment, comparisons have to be made between efficiency of new offerings and existing tools, how new software can be implemented in already successful processes and whether personnel is in place with the knowledge of applying the technology. The goal must be that productivity gains outweigh the costs of digitized manufacturing. SMT
Julian Renz is a Heidenhain product specialist, TNC Controls.
