Introduction

Artificial intelligence (AI) and autonomy have been permeating our daily lives for some time.  Consumers use a plethora of AI-powered virtual assistants like Alexa, Cortana, and SIRI to schedule events, send messages, set reminders, and perform routine tasks.  If you have ever used a GPS app like Google Maps, then you are tapping into the power of AI and autonomy. Google Maps uses machine learning with various data sources like location, historical traffic patterns, and size of the road to assess and predict your best route. It also uses real-time feedback from users and other data sources to reevaluate your route, and if necessary, suggest alternatives. Similarly, industry continues to adopt AI and autonomy. This trend is evident in the progress in areas like robotics, autonomous mining vehicles, offshore subsea operations, and so on. 

In the industrial sector, digital technologies are being used to create smart manufacturing and production facilities. These facilities can take advantage of cognitive technologies to enable adaptive and dynamic responses to a variety of market, equipment, or production process changes. Cognitive technologies can be used to develop self-healing and self-optimizing assets and allow for autonomy across a series of devices, equipment, and business and production systems. These capabilities are emerging from the transition from industrial automation to industrial autonomy (IA2IA). 

In the process industry, you can use industrial autonomy to guide and help operate a unit, improve quality, predict abnormal process conditions, and perform predicative asset management. A process unit running autonomously would be able to detect the plant operating state and make or suggest appropriate set point adjustments to optimize business goals in that state.  For example, a polymer plant can make a grade change that minimizes transition times and reduces waste. The same is true for a refinery making a crude switch. 

Augmenting Human Capability

Industrial autonomy is different from industrial automation. Autonomy means to be independent or to be able to control or govern oneself. This is different from automation, which performs a sequence of preprogramed tasks that require human oversight and intervention between tasks. 

Most people think industrial autonomy means eliminating jobs. That is not the case; many companies are looking toward autonomy to make their employees more productive. In addition, numerous companies are losing expertise, having difficulty attracting new workers, or just do not want to expose workers to dangerous or hazardous environments. This is where autonomous systems can help. They can reduce some physical tasks and significantly augment human decision making. In some instances, autonomy will free people to perform higher-level optimization functions. 

People will work alongside these autonomous components. They will need to understand what the autonomous systems are doing and make sure they are taking the correct and appropriate action or recommendations. The first step might be to use an autonomous system as a digital twin and perform “what-if” analysis. The next step might be to use the autonomous system in an advisory capacity where the operator determines if the recommendation is appropriate and, after implementing it, evaluates its performance. Finally, the system could be put online, and the operator could monitor the system and only intervene by exception. Users should think of autonomy as the GPS for plant operations!

The benefits of autonomy are numerous and include:

• Reducing the amount of time collecting and analyzing data

• Reducing human errors and enabling greater productivity

• Allowing workers to focus on higher value activities like advanced problem solving, analyzing performance, and optimizing processes

• Increasing asset availability with predictive maintenance

• Using data and analytics proactively to predict and solve issues before they cause problems

• Proactively adjusting set points to keep the plant running optimally under changing operating states and objectives

• Enabling automatic startup, transition, and shutdown

• Removing people from hazardous conditions

Where to Start

The steps needed to take advantage of IA2IA will depend on the type of facility and the level of automation. For instance, a greenfield facility can be designed from the start to have higher levels of autonomy. For brownfield facilities, industrial autonomy will likely be incremental with the adoption of autonomous components that accomplish a specific task or an individual function. This could be in the form of an AI algorithm that predicts asset failure and guides repair, detects potential anomalous operating conditions, or determines optimal operating set points. High levels of autonomy could perform closed-loop control actions like opening and closing valves. The autonomous components could be combined and orchestrated to achieve higher levels of autonomy that include other production and value chain functions. 

Initially, autonomy will be focused on repetitive, hazardous, and difficult or error prone activates. For some applications, it may be necessary to convert manual field operations to automatic or perform those tasks with drones and robots. Equipment and processing capabilities might need to be bolstered with additional sensors.  Some typical focus areas include predictive maintenance, quality monitoring, process anomaly prediction, and eventually process control. 

Currently, we are seeing many of our customers trying to achieve unattended remote operations utilizing integrated operations centers (IOCs). This is particularly true for greenfield facilities that can be designed with industrial autonomy in mind to minimize or eliminate manual tasks. This is important for remote or hazardous facilities where it is desirable to remove humans from the site. It also has the advantage that it can significantly reduce initial capital costs. Facilities can be designed to optimize process and equipment reliability by removing points of failure and include additional sensors and cognitive systems to monitor equipment and process conditions to predict anomalies. In addition, it is possible to design tasks to optimize access for robots to take samples, perform inspections, and even conduct routine maintenance.  

The future achieved through IA2IA

There are many benefits to achieving some level of autonomy in the process industry.  Industrial autonomy will enable companies to develop new and enhanced capabilities in all areas, including production, planning and scheduling, distribution and supply chain management, engineering, field operations, and maintenance. Industrial autonomy will provide higher levels of productivity, flexibility, efficiency, reliability and profitability. It will reduce or eliminate human error, provide uninterrupted operations, and remove people from remote or hazardous environments.  Industrial autonomy will help companies achieve their goals to centrally monitor assets and, in some instances, conduct unmanned remote operations. 

Click here to download our eBook and discover more about Yokogawa’s DX approach