More and more businesses are looking to improve the autonomy of their mill or plant operations. An autonomous system in an autonomous mill or plant can monitor its own performance, which has various advantages such as enhanced safety and efficiency, lower costs and environmental effects, and less need for human involvement. However, even with autonomous and optimized operations, humans will continue to play a vital role, one that will focus more on supervising, controlling exceptions and ensuring that diverse process areas work well together.
Security, Dependability and Performance
Because the path to autonomous operations is a journey, shifting responsibilities will be gradual. And, in all autonomous operations, people will still play important roles in both tasks and incident remediation. Technological advances that support continuing employee needs in the short, medium and long term are critical. Two, in particular, are of key importance.
The first concerns alarms and process security. As mode-based alerts for units have developed, alarm administration, records and analysis have greatly evolved. Cross-unit or transitional alarm settings, however, are rarely employed.
Meanwhile, many standards continue to address functional safety as well as safety-instrument systems. It is necessary to develop more robust operations by increasing closed-loop verification and identifying the need for safety element remediation, including sensors, valves and overall logic. Capturing such data and integrating it with the process control domain will be crucial to achieving more autonomous operations, whether it's valve and sensor verification or outcome notification.
The monitoring of process and equipment health and performance is the second area where considerable improvements could be made. Pumps, heat exchangers, compressors and filters, for example, have recently played a larger role in improving facility maintenance and process performance.
While individual pieces of equipment's health and performance are better understood and monitored today, the emphasis remains on single assets. The next stage in engineering oversight is to determine where the temperature of the entering fluid from an upstream exchanger flags an event to the pump and where a shortfall in the pump discharge head alerts a downstream reactor. These cross-asset discussions will ultimately determine autonomous engineering surveillance operations.
Remote and Resilient
As operator operations in the field are reduced, control center locations can be relocated away from the plant site. By transferring personnel from potentially hazardous environments and allowing them to focus on performance improvement rather than data collection and manual analysis, this enhances safety, productivity, and efficiency.
Even project activities, such as changes to control, safety, and security infrastructure, are increasingly being managed remotely, with minimal plant impact until the point of cut-over. This remote method is possible due to the design being decoupled from the real equipment and the use of universal input/output (I/O) technology, which allows control cabinets and channels to be configured for any I/O function.
Remote abilities can also improve resilience. Remote monitoring and troubleshooting for control and safety system infrastructure can improve monitoring, speed up troubleshooting, and reduce the need for site visits. All of this remote assistance can be implemented with great levels of security.
At the same time, autonomous operations will necessitate durable plant infrastructure. Failures will still occur, but the system or process will continue to function normally and automatically recover.
On-site resilience has always relied on duplicate controllers taking over in the event of a failure. A problem triggers the backup. A second problem, however, will still cause an outage. True resilience requires the operation to continue even in the face of repeated failures until the process is exhausted.
This resilience can be achieved by installing multiple process controllers in a distributed mesh, similar to a controller data center. Control methods are no longer limited to a single physical controller. Instead, they are put into a network of controllers that distribute control to wherever capacity is available.
Of course, this is merely the first step toward autonomous operations, which will appear in a variety of settings and across a wide range of activities and disciplines. Many, if not the majority, will find the road arduous. The journey, however, will be worthwhile, and firms will benefit from each step along the way.
Future of Autonomous Operations
Even on a modest scale, achieving autonomy in the process industries can provide benefits. Increased automation and domain application integration, for example, will decrease human error, offer continuous operations, and remove workers from remote or hazardous settings.
In the near to medium term, fully autonomous plants are improbable, but expect some functions to become autonomous to some extent depending on the application, demands, and cost-benefit ratio.
Most process firms operate at or below the Automated level of autonomy, but those at the Semi-Autonomous level may remain there for a long time even as technology, trust and workforce dynamics change. Human interaction and decision-making will continue to be important in many situations, but they will be further reinforced by industrial autonomy as plant workers learn to coexist with these important and increasingly necessary autonomous technologies.
