Showing papers presented at "Analysis, Design, and Evaluation of Human-Machine Systems in 2007"

Smart collaboration between humans and machines based on mutual understanding

Abstract: To improve the safety and comfort of a human-machine system, the machine needs to 'know,' in a real time manner, the human operator in the system The machine's assistance to the human can be fine tuned if the machine is able to sense the human's state and intent Related to this point, this paper discusses issues of human trust in automation, automation surprises, responsibility and authority Examples are given of a driver assistance system for advanced automobile

Combining UML with IEC 61131-3 languages to preserve the usability of graphical notations in the software development of complex automation systems

Abstract: The visual presentation of the software standard IEC 61131-3 for embedded automation systems can hardly impart the structures of complex solutions. The UML as an alternative presentation from conventional software is confusing the engineers of this domain, as it contains many diagrams and notation details, which are designed for other purposes. The UML for Process Automation (UML-PA) offers a customized subset of the UML, which simplifies the graphical structure of the UML. Ambiguous notations elements of the UML are supplemented or cleared. UML-PA refers to IEC 1131-3 behavior specifications to offer skilled familiar notations within the framework of new structures.

Gerico: a human centred eco-driving system

Abstract: This paper presents the evaluation of a new onboard system designed to reduce fuel consumption and CO 2 production. The GERICO system enables the driver to adopt the best driving behaviour, smooth speed and good gear management. We carried out 80 tests (with training and GERICO) with 40 drivers in the Toulouse area. The main results show that training contributes to a 11% consumption reduction whereas using the GERICO system contributes to a 16% consumption reduction. Drivers consider that the system is useful and a good aid to fight against bad habits and reduce fuel consumption.

Understanding human factors in long-distance vehicle operation

Abstract: In this paper, we propose a driver model and a method to analyze driving actions using the model. Driving actions depend mainly on context, i.e. traffic environment and drivers' states. We have developed a set of tools to analyze driving actions and evaluated the effects of traffic environment and the drivers' states on driving actions.

Design and evaluation of a teleoperation interface for heterogeneous human-robot teams

Abstract: Mobile robots can significantly support human teams in dangerous environments in order to reduce risks. Joint teams of humans and robots take advantage of complementary capabilities in navigation, search and planning. Remote coordination of such teams needs efficient teleoperation interfaces. Mixed reality approaches for remote coordination for human-robot teams are investigated. Evaluation results are presented from first experiments.

Development of an intelligent artificial arm

Abstract: The robotics technology is expected to play an important role not only in industries, but also in medicine and welfare. An EMG-controlled artificial arm is one of the most important robotic systems for persons who lost their original arm by accident or disease. This paper presents a concept of the 3DOF intelligent artificial arm that activates the elbow flexion/extension motion, forearm pronation/supination motion, and hand open/close motion based on EMG signals of the user's remained active muscles.

Intelligent task-oriented user interfaces in production environments

Abstract: The impact of user interface (UI) quality has grown in software system engineering. This importance will grow further with upcoming new paradigms such as Ambient Intelligence or Ubiquitous Computing. These paradigms confront the production industry with a new diversity of usage situations. In this paper, we will show the adaptation of a task-oriented, model-based useware engineering process to future paradigms by extending existing models and shifting the generation of the UI from development time to run-time. This integrated engineering process considers several user groups' tasks and UI structure preferences in a common use model described in a system-independent language.

An attempt for conceptual framework for human-machine cooperation

Abstract: As technical processes become more and more complex and the need to support human operators in supervision or control tasks becomes increasingly crucial, the nature of the interactions between human operators and Decision Support Systems (DSS) tends towards Human-Machine cooperation in which the DSS facilitates a partnership with the human operators. In this paper, we first recall a definition for cooperation from the field of psychology: Two agents are cooperating if 1) each one strives towards goals and can interfere with the other (e.g., in terms of goals, resources, procedures) and 2) each agent tries to detect and process such interferences to make the other's activities easier . This definition, which was originally intended to describe human-human cooperation, can be extended to Human-Machine Cooperation if adjustments are made to compensate for the limitations of the machine's capabilities. We then present the four basic elements needed to design a cooperative Decision Support System: • sufficient know-how for solving problems autonomously • know-how-to-cooperate ability • an adequate organizational structure that integrates human and machine, and • a need-to-cooperate . Key studies in engineering have focused on know-how; others have concentrated on the ability to know-how-to-cooperate; still others, on the organizational structure of the Human-Machine partnership and the conditions that motivate human and artificial agents to cooperate to accomplish a task. Using a multi-disciplinary approach that associates research in cognitive psychology and human engineering, we bring these 4 basic elements together in a methodology for designing Human-Machine cooperative systems. This methodology is the result of more than twelve years of experiments at LAMIH in several fields of application, mainly Air Traffic Control and Telecommunication networks.

Evaluation of UML in Process Automation- Results of an Experimental Approach

Abstract: This paper reports on a series of lab experiments to evaluate the application of modelling notations in automation and control engineering. The focus of this evaluation was on two aspects: 1) analyzing and modelling the problem or application and 2) implementing its solution into a control program in IEC 61131-3. The evaluated notations are the Unified Modeling Language (UML) and the Idiomatic Control Language (ICL): UML as nearly standard in computer science and ICL as powerful modelling notation from the distributed control systems world (DCS).

Precision and repeatability analysis of optotrak certus as a tool for gait analysis

Abstract: This study introduces a new method of determining precision and repeatability of Optotrak Certus, a tool for gait analysis using a three dimensional (3d) robot. Using 3d robot as a reference of standard, the accuracy and precision of the system in terms of angle and distance were quantitated. Angle and distance among markers showed good agreement between measurements, and comparable measures of precision were reported. Experimental results containing CMC, CV and SD values for variation of angle, volume, circular movement and speed demonstrated, provided a measure of precision and repeatability; hence the repeatability was satisfactory. The investigation is ethical and practical in measurement analysis.

BCD model for human state identification

Abstract: The paper presents an attempt to use the so-called BCD model (i.e., Benefit/Cost/Deficit model) in order to identify the human state. Different human states are taken into account from the automation and mechanical viewpoints to the psychological and physiological ones. Various signals identifying these human states can be analysed firstly alone, and then together in order to identify Benefit, Cost and Deficit parameters of a situation, here a driving traffic situation, and to analyse the impact of human state stability on safety driving.

Potential and challenges of stereo augmented reality for mobile robot teleoperation

Abstract: In many mobile robot applications a human operator is still required to control the robot. The design of the user interface is the major element affecting the performance and efficiency of the task execution in the teleoperation scenario. This work investigates an approach to visualize all information gathered by the mobile robot in an integrated fashion by applying stereo augmented reality. Initial experiences and results with this type of interface for navigation tasks are presented.

Situation and intent recognition for risk finding and avoidance

Abstract: This paper gives a brief overview of the three-year research project, 'Situation and intent recognition for risk finding and avoidance.' The project was launched in 2004 with the support of the Government of Japan, for developing proactive safety technology to detect mismatches among traffic environment, driver's recognition of it, and intent of the driver so that the driver may be provided with an appropriate assist functions in a situation-adaptive and context-dependent manner. This paper gives an introduction to the papers delivered in the organized session, 'Context-dependent driver support for automobile safety.'

Vision-based gaze estimation method for stereoscopic display systems

Abstract: In this paper, we propose a new near infra-red (IR) vision-based gaze tracking method, which is not affected by facial movements in stereoscopic virtual reality (VR) systems. Our proposed method has following five novelties compared to previous works. (i) By using our gaze tracker instead of a conventional input device, users felt more interested and immersed in the stereoscopic VR system, (ii) by using an IR light passing filter and an IR illuminator, we were able to implement a robust gaze tracker irrespective of environmental lighting conditions, (iii) we used the simple 2D based eye and face gaze estimation method without considering the complicated 3D structure and motion of human eye and face, which could reduce the system complexities, (iv) to prevent the unintended movement of gaze position by natural eye blinking, we discriminated the eye blinking by measuring pupil size. And this information was used for button clicking and mode toggling based on the analysis of user's fatigue. That is, the measured fatigue information of user was used for the mode toggling between gaze tracking mode and the conventional mouse input mode, (v) the proposed method allowed the natural movement of user's head by tracking the IR rhombus-shaped pattern, which was not affected by the change of environmental lighting. Experimental results showed that the gaze estimation accuracy of the proposed method was 0.98 degrees. In addition, the subjective tests showed that the VR application using the proposed method was more interesting and immersive to users than those using conventional input devices.

Model-based development of user interfaces a new paradigm in useware engineering

Abstract: Ubiquitous computing, ambient intelligence, and all of the related interaction paradigms for our computerized future require the use of modern, development methods. As the complexity and spread of microprocessor based devices increases, so does the need for user oriented adaptation to intelligent environments to include customization of the user interfaces for an uncontrollable number of user types, operating situations, and user preferences. Usability experts and interaction designers must reorient themselves now for the use of model-based methods in the development of user interfaces.

Functional modeling and constraint management in command and control: two microworld studies

Abstract: Command and control settings are increasing in complexity, and are likely to continue to do so with the envisioned increase in communication and cooperation between interdependent networked units. This paper describes two microworld studies, with a focus on a method for the management of constraints in network-based command and control. The method enables the recognition, modeling, and analysis of constraints, combining goals-means task analysis, functional modeling, essential variables analysis, and generation of constraint-based recommendations for the design of tasks and interfaces. The first microworld study utilizes an emergency management task, the second an adversarial command and control task.

The tactobook: a portable electronic tactile reading device for the blind

Abstract: This paper introduces a novel concept on reading assistive technologies for the blind/visually impaired: the TactoBook , a system that is able to translate entire text books to Braille code and to reproduce them in portable electronic touch stimulation devices. The TactoBook consists of a computer-based system that translates fast and automatically any text book into Braille. The Braille version of the book is then stored in a USB memory drive which is later inserted and reproduced in a compact, lightweight and highly-portable tactile display. The user freely explores the information displayed and is able to read comfortably the book without carrying the burdensome tactile print one.

Multimodal human-machine interface and user cognitive load measurement

Abstract: Multimodal user interface (MMUI) is an emerging technology that aims at providing a more intuitive and natural way for people to operate and control a computer or a machine. MMUI allows users to control a computer using various input modalities, including speech, touch, gestures and hand-writing. It has potential to minimise user's cognitive load when performing complex tasks. In this paper we present our work in building an MMUI research platform for intelligent transport system applications, and our attempt to evaluate a user's cognitive load based on analysis of his or her multimodal behaviours and physiological measurement.

Simulating surface roughness in virtual environments by vibro-tactile feedback

Abstract: One of the most important objectives of virtual environments is to provide an immersive rendering of real-world objects It is assumed that human performance will be best if the operator is able to see, to touch, and to manipulate objects as he would do in everyday life The main emphasis of this work is put on the immersive design of haptic feedback; thereby particularly the simulation of surface roughness in virtual scenarios is regarded in more detail Below, both the actuator concept and the rendering approach of such a system are proposed (see Section 1) Afterwards the results of an experimental evaluation are presented (see Section 2, 3) and its impact on further research activities is discussed (see Section 4)

Detection of cognitive distraction via a drier-adaptive sensor fusion

Abstract: This paper analyzes effects of performing a cognitive subtask during driving and develops methods to detect cognitive distraction. The results suggest that how to integrate each single variable judgment should be determined depending on to what extent the false alarm rate is allowed. By conducting a discriminant analysis, this paper also suggested that fusion of multiple dimensional sensor data can improve the reliability of the detection of cognitive distraction.

Evaluating the mariner's workload using the bridge simulator

Abstract: Many maritime accidents are caused by human error. Human error is mostly affected by the state of the mariner's physiological and psychological aspects. For example, excessive stress creates error, lapses of attention cause the mariner to overlook or fail to notice problems. To clarify how physiological and psychological states affect human involve many problems to be solved.

Using glyphs to facilitate transition from hierarchical selection to gesturing

Abstract: The successive strokes made in multi-stroke marking menus form gestures. This paper reports on an experimental study that examines how the use of glyphs affects performance in marking menu selection. Glyph familiarity is manipulated by comparing Chinese style and Mark style glyphs across Chinese and non-Chinese users. Performance benefits are found for familiar and memorable glyphs. The results are discussed in terms of their implications for the design of marking menus and associated gestures and for the role of glyphs in facilitating the transition from novice to expert performance in menu selection.

Automobile vehicle ingress/egress movement strategies adopted by elderly and/or, hip or knee, prosthesis having persons

Abstract: This paper presents the automobile ingress/egress movement strategies adopted by 41 test subjects from a mixed population of able-bodied, elderly and/or disabled drivers when entering/exiting four stripped vehicles representing 90% of the automobile types presently on the market. Two general movement families were identified, those starting from a standing position on one foot or on two feet. Eight ingress/egress movement strategies were identified (5 ingress and 3 egress). This paper describes these strategies in terms of the population performing them and the vehicle tested. It appears that no one strategy is used systematically by a single person, since subjects tend to adapt their strategy to the tested vehicle.

Modeling robot movement in a virtual environment

Abstract: This research looked at the use of robots in place of humans in urban search and rescue (USAR) missions. The success of USAR missions often rests on the ability of the human to control the robot. This study evaluated the following three control modes for USAR missioning: gamepad, keyboard, and mouse. Task completion time and error count were the performance measures used. The results suggest that for robot maneuvering which requires coordinated control of both robot movement and directing the camera, the use of a game pad achieves higher performance than the mouse or keyboard. On the other hand, for the control of robot arm movement, all three control modes perform equally well.

Accident based requirements analysis for advanced driver assistance systems

Abstract: Besides increasing the comfort for drivers by the use of automotive assistance systems like cruise control or navigation systems new advanced driver assistance systems like lane departure warning or the emergency brake assistance intend to increase road traffic safety. Further the increase of safety goes along with mitigating risks, which for road users are expressed by the possibility and the severity of accidents. The presented approach in this paper shows a methodical way of learning from accident characteristics and statistics recorded in databases in order to set up functional requirements for new driver assistance systems that are able to improve the road safety by efficiently assisting the driver in critical situations on one hand and to set the main focus on functional assistance systems for situations with the highest accident rates on the other hand. This will help to evaluate existing systems and to develop advanced and safety efficient driver assistance systems increasing road safety

From overcoming limitations to securing safety: defining the new mission of human-machine systems

Abstract: Beginning with the launch of human factors engineering in 1945, but actually with intellectual roots that go back to at least the school of Scientific Management, the design of human-machine systems (HMS) has been focused on overcoming human limitations and constraining human performance variability. This approach is, however, only warranted if the work environment is simple enough to be tractable. While this may have been the case 50 years ago, it is no longer so. Most work systems today are intractable, due to tight couplings and non-linear interactions. Under these conditions human performance variability becomes an asset rather than a liability, and the goal of HMS design should therefore be to enhance the capabilities of humans rather than to constrain them.

iFace: A FACIAL EXPRESSION TRAINING SYSTEM

Abstract: The human's facial expression plays an important role as media that visually transmit feelings and the intention. In this study, it aims at a support of the effective process for facial expression training to achieve the target expression using computer, as a self training system. And, an interface for users to select a target facial expression and a whole development of an effective expression training system is proposed, as a first step toward an effective facial expression training system.

Human complementary menu design for mobile phones

Abstract: This study proposes user control and system support as means to facilitate system-driven and user-driven interface adaptations of mobile phones. A hypothetical prototype of mobile phone menu with an additional collection of quickly accessible items, which is called 'hotlist,' was implemented. An experiment was conducted for the evaluation of five different menu types: traditional (T), adaptable (B), adaptable with system support (Bs), adaptive (V), and adaptive with user control (Vc). The results showed that providing the system support to the adaptable menu not only increased the users' perception of the efficiency of selection, but also reduced the menu adaptation time.

A core task analysis based work support and training tool for control engineers

Abstract: This paper presents a study in which a concept of a work support and training tool for control engineers was developed The aim of the tool is to provide control engineers means for developing professional skills and to promote collaboration and sharing of knowledge among the community of engineers First, an interview study was conducted to gain understanding of the characteristics and demands of control engineering work After identifying user requirements the concept of the tool was defined Finally, a prototype was implemented using web technologies and emerging standards of computer supported learning

Lane-change behavior modeling of long-haul truck driver using driving database

Abstract: Driving behavior data measuring systems are constructed to collect many driving behavior data in long-haul driving. Such systems are installed in several heavy-duty trucks for transportation business. Many long-haul driving behavior data of several drivers on expressways are collected for a long time. These driving behavior data are integrated into a database. Using this database, typical driving behavior patterns, such as lane changes, are extracted and their variation is analyzed statistically.