Jan-Patrick Osterloh

Modeling Pilot and Driver Behavior for Human Error Simulation

In order to reduce human errors in the interaction with in safety critical assistance systems it is crucial to consequently include the characteristics of the human operator already in the early phases of the design process. In this paper we present a cognitive architecture for simulating man-machine interaction in the aeronautics and automotive domain. Though both domains have their own characteristics we think that it is possible to apply the same core architecture to support pilot as well driver centered design of assistance systems. This text shows how phenomena relevant in the automobile or aviation environment can be integrated in the same cognitive architecture.

Simulating Perceptive Processes of Pilots to Support System Design

In this paper we present an approach towards supporting the ergonomic design of aircraft cockpits by predicting the probability that pilots might miss relevant information due to routine learning effects combined with non-adequate placement of display instruments. The approach is based on an executable cognitive pilot model. We focus on the cognitive interaction between (1) rule-based processing of flight procedures, (2) the pilots mental model of the current situation and (3) pilots attention. The cognitive model is coupled with a formal cockpit design to simulate human-machine interaction during flight procedures. As an example we analyze the perception of automatic flight mode changes.

Selecting Human Error Types for Cognitive Modelling and Simulation

This paper presents a method that has enabled us to make a selection of error types and error production mechanisms relevant to the HUMAN European project, and discusses the reasons underlying those choices. We claim that this method has the advantage that it is very exhaustive in determining the relevant error types and error production mechanisms, and that the final objects are selected according to explicit requirements, without missing relevant error types and error production mechanisms.

Cognitive modelling of pilot errors and error recovery in flight management tasks

This paper presents a cognitive modelling approach to predict pilot errors and error recovery during the interaction with aircraft cockpit systems. The model allows execution of flight procedures in a virtual simulation environment and production of simulation traces. We present traces for the interaction with a future Flight Management System that show in detail the dependencies of two cognitive error production mechanisms that are integrated in the model: Learned Carelessness and Cognitive Lockup. The traces provide a basis for later comparison with human data in order to validate the model. The ultimate goal of the work is to apply the model within a method for the analysis of human errors to support human centred design of cockpit systems. As an example we analyze the perception of automatic flight mode changes.

Automated UI evaluation based on a cognitive architecture and UsiXML

In this paper, we will present a method for automated UI evaluation. Based on a formal UI description in UsiXML, the cognitive architecture CASCaS will be used to predict human performance on the UI, in terms of task execution time, workload and possible human errors. In addition, the UsabilityAdviser tool can be used to check the UI description against a set of usability rules. This approach fits well into the human performance and error analysis proposed in the European project HUMAN, where virtual testers (CASCaS) are used to evaluate assistant systems and their HMI. A first step for realizing this approach has been made by implementing a 3D rendering engine for UsiXML.

Modelling and Validating Pilots’ Visual Attention Allocation During the Interaction with an Advanced Flight Management System

This paper presents the results of our analysis of human pilot behaviour and of a cognitive pilot model. We have performed experiments in a flight simulator including a new 4D flight management system (FMS) in order to gather information about the interaction of human pilots with the new FMS and to validate the performance of the cognitive pilot model. This paper focuses on the visual attention allocation of human pilots and on the validation of the visual perception component of the pilot model.

Towards Model-Based AHMI Automatic Evaluation

Aircraft cockpit system design is an activity with several challenges, particularly when new technologies break with previous user experience. This is the case with the design of the advanced human machine interface (AHMI), used for controlling the Advanced Flight Management System (AFMS), which has been developed by the German Aerospace Center (DLR). Studying this new User Interface (UI) requires a structured approach to evaluate and validate AHMI designs. In this paper, we introduce a model-based development process for AHMI development, based on our research in the EUs 7th framework project “Human”. The first goal is to rely on this structured approach to perform automatic evaluation of the User Interface.

Distributed Power Generation: Requirements and Recommendations for an ICT Architecture

Contemporary power distribution faces various new challenges. Most of those challenges have a strong impact on the ICT-structure required and on system architecture. This contribution briefly introduces changes and requirements imposed, both for trading and distribution of power. On this basis, alternatives for ICT architectures are discussed, recommendations are made on implementation choices for a meaningful sustainable solution, and communication and security challenges are addressed.

Monte Carlo Methods for Real-Time Driver Workload Estimation Using a Cognitive Architecture

Human-machine interaction gets more and more cooperative in the sense that machines execute many automated tasks and cooperate with the human operator, who also performs tasks. Often some tasks can be executed by both, like a car that can autonomously keep the lane or is steered actively by the driver. This enables the human machine system to dynamically adapt the task sharing between machine and operator in order to optimally balance the workload for the human operator. A prerequisite for this is the ability to assess the workload of the operator in real-time in an unobtrusive way. We present two Monte Carlo methods for estimating workload of a driver in real-time, based on a driver model developed in a cognitive architecture. The first method that we present is a simple Monte Carlo simulation that gets as input the information that the driver can perceive, but does not take the actions of the driver into account. We evaluate it based on a driving simulator study and compare the workload estimates with functional near-infrared spectroscopy (fNIRS) data recorded during the study. Afterwards the shortcomings of the simple approach are discussed and an improved version based on a particle filter is described that takes the driver’s action into account.

Model-Based Pilot Training Syllabus

Technical progress takes place rapidly in every industry and so it does in aviation. Nowadays formerly acquired flying skills of a pilot doing transition training from one aircraft to another are completely ignored. Thus, the trainee is faced with a lot of training on already existing knowledge, without being properly advised to the fine differences and specific characteristics of the new aircraft. Our idea for solving this deficit is a tool that compares the knowledge a pilot already has to the knowledge needed for the new aircraft type. To do so, we modeled the whole set of tasks for flying an aircraft, down to the lowest level of key-strokes, for the different aircraft types. Based on the models, our algorithm finds the differences in the task models and creates syllabi which properly incur to the sticking points of these differences and allow recognizing what kind of learning transfers will occur.