Rajagopalan Srinivasan

Eye gaze movement studies of control room operators: A novel approach to improve process safety

Abstract Process industries continue to suffer from accidents despite significant regulatory intervention since the mid-1980s. Human error is widely considered to be the major cause for most accidents today. Detailed analysis of various incidents indicates that reduced staffing levels in control rooms and inadequate operator training with complex automation strategies as common reasons for human errors. Therefore, there is a need to develop deeper understanding of human errors as well as strategies to prevent them. However, similar to hardware failures, traditionally human error has been quantified using likelihood approaches; this viewpoint abnegates the role of the cognitive abilities of the operators. Recent studies in other safety critical domains (aviation, health-care) show that operators level of situation awareness as inferred by eye tracking is a good online indicator of human error. In this work, a novel attempt is made to understand the behavior of the operator in a typical chemical plant control room using the information obtained from eye tracker. Experimental studies conducted on 72 participants reveal that fixation patterns contain signatures about the operators learning and awareness at various situations. Implications of these findings on human error in process plant operations them are discussed.

A Novel Experimental Strategy for Validating Human Failure Probabilities in Risk Assessment

Abstract Human error is a significant contributor to process accidents today. Human failure probabilities are mostly based on the judgment of experts; this makes conforming their validity difficult. In this paper we propose a novel experimental strategy for validation of human failure probabilities. Our strategy is based on the time taken by human participants to bring a plant to a normal state, in response to an abnormal event (like disturbance in flow rates or temperature). The inability of human participants to bring the plant within normal limits is considered to be human error (failure). The response times taken by human participants give the measured human response-time probability distribution. The response-time distribution enables us to validate the human response times and, consequently, human failure probabilities. We report the results obtained from a study of 72 participants controlling an industrially important process.

Quantifying situation awareness of control room operators using eye-gaze behavior

Abstract In an attempt to improve process safety, today’s plants deploy sophisticated automation and control strategies. Despite these, accidents continue to occur. Statistics indicate that human error is the predominant contributor to accidents today. Traditionally, human error is only considered during process hazard analysis. However, this discounts the role of operators in abnormal situation management. Recently, with the goal to develop proactive strategies to prevent human error, we utilized eye tracking to understand the situation awareness of control room operators. Our previous studies reveal the existence of specific eye gaze patterns that reveal operators’ cognitive processes. This paper further develops this cognitive engineering based approach and proposes novel quantitative measures of operators’ situation awareness. The proposed measures are based on eye gaze dynamics and have been evaluated using experimental studies. Results demonstrate that the proposed measures reliably identify the situation awareness of the participants during various phases of abnormal situation management.

Non-intrusive Appliance Load Monitoring for Electrical Energy Systems Simulation and Analysis – A case study in India

Abstract Residential customers play an important role in total power consumption with a highly varying demand profile depending upon climatic conditions. With recent technological advancements, it is possible to deploy Renewable Energy Sources (RES) at customer end to support their power demand. To take advantage of this opportunity, it is important to understand the load profile of residential users. Also, deployment of RES at customer end will lead to a more decentralized power grid adding to the complexity of the system. Modelling and analysis of the resulting system is therefore necessary for optimal electric energy utilization. Bottom up approaches estimates the demand profile of customers which is then utilized to obtain consumption patterns at distribution and generation levels. Demand profile of residential users could be estimated using a technique known as Non-intrusive Load Monitoring (NILM). NILM disaggregates the aggregate power consumption measured at utility entry point to identify the operational states of individual appliance. In this work, we propose an NILM technique which uses Neural Network (NN) along with Hidden Markov Model (HMM) to detect the operating state of an appliance. The appliance level consumption obtained using this approach for a community can be then utilized for developing bottom up models for energy system simulation.

Cognitive Engineering for Process Safety: Effective Training for Process Operators Using Eye Gaze Patterns

Abstract Control room operator errors are considered a major reason for accidents in the process industries. Inadequate training for operators is often considered one of the major causes of such accidents. A research report aimed at improved crew resource management in the Oil and Gas industry suggested that a systematic analysis for trends in human factors causes of accidents is important for development of operator training program. In addition process safety experts suggest training program with emphasis on operator decision making process for operator learning enhancement. Eye tracker based studies have been devoted to learning enhancement in various safety critical domains. In our previous research, we employed an eye tracker and observed the existence of characteristic gaze patterns for successful chemical plant operators. In this work, we use these eye gaze patterns along with results from a game based learning approach to develop a training methodology to impart process knowledge to plant operators. Our experimental studies using an ethanol plant simulator on 14 graduate level participants reveals the presence of characteristic fixation pattern that contain information about learning advancements.

Integrating Control and Scheduling based on Real-Time Detection of Divergence

Abstract Scheduling and control have been long recognized as the two critical building blocks in many manufacturing execution systems. Operating at the interface between the supply chain and the process, the scheduler generates a detailed schedule that has to be executed by the process so as to meet the demands originating from the supply chain. Given the tight interactions between the two, there has been wide interest in integrating control and scheduling. A variety of methods ranging from monolithic integration into one large integrated problem, to hierarchical cooperative approaches have been proposed in literature. In this paper, we propose a novel approach to the integration problem. Our key insight is that disturbances which occur post the generation of the original schedule, trigger a divergence between operational targets defined by the schedule and its execution. If left uncorrected, the disturbances will propagate between the process and the supply chain. A timely response could eliminate or minimize such effects. Recognizing this, we propose a novel framework for integrating scheduling and control that detects in real-time when a divergence occurs between the original schedule and its execution in the process, identifies the root-cause(s) of the divergence (i.e. the disturbance), and triggers a suitable response from the scheduler and the process so as to nullify or minimize its effect. In this paper, we will describe the proposed approach and illustrate it using an industrially motivated case study.

Structural Similarities and Differences between Smart Grids and Process Industry Supply Chains: India Case Study

Abstract Process industry supply chains have been an active area of research for decades with advances towards decentralized management and development of closed loop resilience. Along similar lines, now electric power systems are moving towards smart grids that allow for decentralized power generation. At the surface level there are many differences between electrical grids and supply chains; for example, there is limited/no storage in case of electric power, hence supply and demand has to be matched at every time instant. However, in the recent times it has been recognized that some of the control and enterprise-wide optimization strategies that are an integral part of supply chain management have analogues in power systems as well. This article seeks to systematically identify the structural similarities and differences between these hitherto unrelated domains. We will illustrate these using a case study of Indian electrical grid.

Dynamic Simulation-Based Assessment of Supply Chain Sustainability

Abstract As sustainability becomes an increasingly important business factor, companies are looking for decision support tools to assess the impacts associated with their manufacturing operations and supply chain activities. Life cycle assessment (LCA) is widely used to measure the environmental consequences and more recently social/societal impacts of a product throughout its life cycle. However, LCA-based assessments are static as they do not consider the dynamics arising from the multitiered structure and the interactions along the supply chain. In this work, we describe a framework integrating dynamic simulation with LCA indicators for sustainability assessment that considers the dynamics in supply chain operations. The advantages of this framework are demonstrated through sustainability assessment scenarios involving changes in product composition, ordering policy, and supplier selection policy in the diaper and detergent supply chains.