In prior installments of this series on industrial mobile robots, we established that an emerging shortage of personnel to operate plant equipment and process units has resulted in a growing need for robots and drones that can travel throughout facilities and take over data collection, inspection, and other tasks.
In addition to the labor shortage, the most common challenges in routine inspections and maintenance activities today include safety risks due to work in hazardous areas and limits in human capabilities in accurately and efficiently acquiring and processing information.
Their ability to reach dangerous areas quickly, easily, and safely enables robots and drones to improve the overall safety of plant operations and maintenance. They also allow the human workforce to focus on value creation tasks, which are far more attractive to prospective employees. (Hereafter, I will mention only robots but with the understanding that, in practically all cases, the comments apply to drones, as well.)
Advanced robot technologies can even help end-users achieve unattended operations, while also augmenting human capabilities to solve the common challenges in daily plant operations and routine inspections.
Providing insightful data, that is, bringing the right data to the right viewers such as key decision makers, is among the most promising robot capabilities. The robots can be equipped with sensors and specifically-designed payloads to meet operational objectives.
However, data integration is always a challenge. Robots can collect a lot of data. Some prospective users have literally asked, “but is it useful? What do we do with it?”
This is not really the best way to look at it. The most advanced users have distinguished themselves by their ability to make the most of the data provided by robots. They specifically define the data collection requirements for each mission, match the payloads, and send the robots on their way.
The data collected by the robots falls into two fundamental categories: operations and maintenance. An example of the former would be a process variable such as a pressure gauge reading. Although the gauge is “disconnected,” the value typically does make its way to the process control system. An operator would manually enter it upon completion of the round. Often, this sort of data is only for recording purposes in the historian.
An example of maintenance information is a vibration measurement, which could indicate that an equipment repair is needed soon. The ultimate destination of much of the maintenance information is an asset management application.
Compared to human operators or technicians making rounds, robots offer the advantages of accuracy and timeliness. Robots eliminate the “human error” that could affect the manual measurements during rounds and data entry afterward. If a robot resides on a wireless network, it can immediately report data recorded during a round. There is no need to wait until the round is completed.
Since the data collected by a robot is destined to asset management applications, historians and process control systems, significant processing is required within the robot to arrive at a data format that is appropriate to those platforms. For instance, when a robot reads a pressure gauge or measures vibration, it converts those data points to the same digital formats that the destination platforms use.
Robots typically activate applications that are based on their missions and payloads. To read gauges, a robot would use an application such as “Plant Image Analyzer” (PIA), which processes image data from plant sites using AI technology. PIA analyzes images of devices such as gauges and converts readings to digital values. It includes a gauge calibration function, which allows the robot to determine the minimum or “zero” value, maximum value, and the engineering units.
In addition to converting data to digital formats, robots can record audio, photos and videos. This sort of information has proven valuable to process operations and maintenance management. Not only can the various formats be viewed and heard on HMI devices and web browsers, but AI applications can offer much deeper analyses.
An example of a specialized AI application is a color degradation diagnosis system, which employs a robot integrated with an AI image analyzer to assess objects such as process pipes for color changes. The sensors capture data on hue and brightness, which AI algorithms interpret to diagnose degradation severity. The system then flags items for maintenance, offering fast and accurate assessments to preserve object quality and value.
In order to transport the data collected by robots to destinations such as asset management applications and process control systems, industrial companies today use specialized platforms, which have evolved from SCADA systems. Modern versions, for example, Yokogawa’s “Collaborative Information (CI) Server,” can reside on-premises or in the Cloud. It integrates the robot fleet management functionality with a broad array of communication interfaces and protocols, which offer compatibility across OT technologies.
Among numerous emerging applications for industrial mobile robots, such capabilities enable a first responder scenario. When the CI Server detects alarms from field devices, a robot can be dispatched to a field location and the real-time field situation can be checked with image data from the robot.
In emergency situations, deploying robots as first responders offers numerous advantages. Those include the support for operators to assess the situation, gather critical data, and execute tasks such as deactivating equipment or assisting in rescue efforts.
When integrated with the control system, a robot can continuously provide real-time feedback to the command center, enabling responders to formulate effective strategies and mitigate safety risks.