Remote machine monitoring offers information with context
October 20, 2011
The evolution of the smartphone has provided the opportunity to improve notification systems to include vital information along with the standard message. Imagine having real-time access to machine and production status from home or the cabin. But like an iceberg, there is more to it than meets the eye, and what you don’t see is far more important than what you do.
Remember the last time you were left in the dark because of a power outage? The minutes seem to drag along like hours. Did you receive an automated message from the utility company’s notification system informing you of the outage? If so, I am sure you would agree that the message was essentially meaningless. You know the power is out. What you want to know is when the power will be turned on so you can return to your normal life.
Unfortunately, notification systems have limited value. An automated message is meaningful to the lineman who repairs the problem. But to you, the message lacked context. Had the message included an estimated repair time, you could have made some informed decisions. For a notification system to have real value, it must include information that puts the message into a context for the recipient.
In other words, we need information, not data.
With the evolution of the smartphone, we have the opportunity to improve notification systems to include vital information along with the standard message. Imagine having real-time access to machine and production status from home or the cabin. The idea sounds very desirable. But, like an iceberg, there is more to it than meets the eye, and what you don’t see is far more important than what you do.
So how does this information make its way from the shop floor to a smartphone? For a production manager to receive a message about stopped production caused by a machine error, three processes have to be in place. First, the CNC machine tool or automated process must be able to generate valid data such as error codes or event triggers. Second, the data must be gathered in a central location where context can be added, creating meaningful information. Third, that central location must securely communicate the information to the production manager.
Regarding data generation from CNC machine tools, data triggers can occur one of two ways. The intramachine method effectively places in the control a kind of “agent” that monitors the status of the machine tool and generates data on predetermined events. Not all machine controls accept a monitoring agent, but those that do can offer significant benefits. The intramachine agent has access to valuable information not available with other methods. Intramachine monitoring requires collaboration among software vendors, the CNC manufacturers that provide the software, and the machine tool builders who provide access to the data they generate.
The other method of gathering machine tool data is with external machine monitoring products. These tools include relays, sensors, and human-machine interfaces. External monitoring is the most common method of gathering machine tool data for two primary reasons. First, there wasn’t a reliable intramachine agent system available until quite recently, and second, external monitoring is flexible, less expensive, and requires less expertise to implement.
The main problem with external machine monitoring is that only limited data is generated by the device. A hard-crash error will generate the same notification as a soft error that can be corrected with a simple reset button. Ancillary information is unavailable, which can reduce the quality of information in the notifications.
A common tool used with external machine monitoring is the human-machine interface (HMI). An HMI is a computer or device that allows an operator to easily communicate conditions or status to the monitoring system. These black boxes are easy to use and can provide more data than passive monitoring devices.
The problem is that HMIs typically require an operator to be engaged in the monitoring process. Any process that involves a person to just push a button to report an event makes recording that event unreliable. Even the best personnel forget to push the button or go around the process from time to time. In short, any HMI monitoring system that is not transparent in the production process introduces incomplete data, and therefore the monitoring process is suspect.
Once generated, data is sent to a central location for processing and passing on to the end user. If the recipient is located within the company walls, the process is very straightforward. The message is simply relayed via messaging, e-mail, or Web service using the existing company network.
Relaying errors while key personnel are nearby does not require a major investment in a notification system. If the notification moves outside of the company walls, however, the complexity of the system rises—but so does its value: Key personnel can be anywhere yet still be in control of production.
As discussed, a simple error message does not contain enough information to allow the recipient to effectively make a decision. Notifications must have meaningful information added so that the recipient understands its impact. For instance, a production manager who receives a CNC error notification will call a maintenance person and ask how long it will take to fix. If that same notification included an error report and projected production schedule, the manager would understand the impact of the notification. Many notifications are minor, so the production manager does not have to take action. However, when a major failure occurs, the manager should be aware of the implications and respond accordingly.
Where does the information come from that makes notifications intelligent? The most common source is the company enterprise resource planning (ERP) or manufacturing execution system (MES), as they are typically in control of production schedules. Other sources like a shop’s distributed numerical control (DNC) systems also will work, but the goal is always the same: Synchronize with the trusted production schedule. This synchronization can be quite complex and may require qualified technicians to perform the work. It also represents a large portion of the success or failure of the project, yet it is often underestimated in terms of value and return on investment.
All monitored CNC equipment and processes are connected to a data collection center, which serves three purposes:
Today most CNC machine tools can be integrated into the company network. Some companies even have implemented notification systems that directly send errors to people outside the company walls. While well-intentioned, it simply isn’t safe to put the CNC machine tools on a network that has direct access to the Internet. The CNC equipment lacks the necessary safeguards to be exposed to the Web. Instead, machine tools should be isolated with a proxy server that will relay data between the CNC equipment and the Internet. Being a dedicated server, it can be set up with the necessary security tools and protocols to ensure machine tool data can be accessed safely online.
Once up and running, the data collection server gathers all signals and notifications from the shop floor, generates intelligent notifications, and forwards them to the proper recipients. Messages can be sent in many ways. E-mail notification remains most popular simply because it is reliable and uses existing e-mail technology found in all companies. Some, though, are choosing to use messaging services that are quick and easy to access via cell phones.
Both of these methods allow the notification system to embed hyperlinks back to the server. The data collection system can generate substantial information surrounding an error, but can’t effectively package it all to send to the end user. For this reason, the hyperlink allows a one-button click that tunnels back to the data collection system to provide great detail about the notification.
Hyperlinks and remote server access have security issues. If a simple push of a button allows someone to see detailed information, what would prevent a competitor or nonauthorized person from seeing the information? Secure identity management must be in place for any company that utilizes remote access of data. This security can take many forms, such as virtual personal networks (VPNs) and security certificates, but it must be taken into account for any system.
This leads us back to smartphone technology. Let’s assume the notification system is properly set up to pass meaningful information from the data collection system to the outside world. With the scenarios described previously, the end user receives secure information about processes within the company walls. But that information is static and he is passive. He can receive only what’s given to him—that e-mail, which remains static, unchanging.
Some applications found in data collection systems can migrate to the smartphone. Now end users simply can ask the status of a machine or process, or ask for a history of all errors in the past 24 hours. In essence, end users create their own information; they become active and the information becomes dynamic.
What impact does this have? People can do their jobs based on knowledge they generate, not what is sent to them. It also broadens the scope of the data collection system from a notification/monitoring system to a production communications server. Because all aspects of the production process are available, the system can be used to learn more about production problems and make quick decisions.
Although the production manager isn’t at work, he isn’t in the dark about production. His smartphone has just notified him of a machine failure and, more important, the fact that critical work will not finish on time. A quick query indicates that all other machines are functioning normally and not running hot jobs. Rescheduling the hot jobs to other machines, the production manager successfully has put the shop floor back on track. A follow-up query later that evening confirms that the critical production has been completed. The system also indicates that all machines are up and operating, easing concerns over extended downtimes. Tomorrow’s production meeting will provide everyone with the details of the event and its impact moving forward.
It isn’t a question of if unplanned events will impede production. The question is when. Thanks to the Internet, cloud computing, and smartphone technology, those unplanned downtimes do not have to create chaos within the production environment.