scispace - formally typeset
Search or ask a question

Showing papers by "Morten Fjeld published in 2005"


23 Nov 2005
TL;DR: For enhanced interaction, keyboard-free system configuration, and internal/ external database access, a graphic user interface (GUI) has been incorporated into the TUI, and three-dimensional rendering has been improved using shadows and related effects, thereby enhancing depth perception.
Abstract: Augmented Chemistry (AC) is an application that utilizes a tangible user interface (TUI) for organic chemistry education. Based on the outcome of an extensive evaluation, we are in the process of extending the AC system. Firstly, for enhanced interaction, keyboard-free system configuration, and internal/ external database (DB) access, a graphic user interface (GUI) has been incorporated into the TUI. Three-dimensional (3D) rendering has also been improved using shadows and related effects, thereby enhancing depth perception. Secondly, AC has been ported to different operating systems and is now compatible with Linux-, Windows-, and Mac OS X based platforms. This enables the use of a wider range of hardware: USB and Firewire (IEEE1394) cameras are now supported. Finally, system capacity to import and visualize molecules from an extensive XML-based DB has been realized. This gives users the ability to download and interact with any molecule up to a certain complexity.

19 citations


Proceedings ArticleDOI
05 Jul 2005
TL;DR: This paper presents a haptic device inviting users to interact with the law of physics, based on an in-house design of a generic force feedback slider (FFS), and realizes a software application simulating a catapult.
Abstract: Physics education often relies on visualization of theoretical laws. While Java animations are widespread, they mostly lack user interaction. We propose a haptic device inviting users to interact with the law of physics. Based on an in-house design - a generic force feedback slider (FFS) - we have realized a software application simulating a catapult. As users interact, they receive both tactile and visual feedback. By calling upon two perceptual channels at a time, here tactile and visual, we assume users may construct their mental model more easily. This paper presents our application, the underlying FFS technology, a user study, future uses, and a discussion.

17 citations


DOI
01 Jan 2005
TL;DR: A detailed software and hardware architecture for a force feedback slider (FFS) is reported on and a simple Application Programming Interface (API) is offered to enable integrated FFS applications to build applications incorporating a FFS.
Abstract: We report on a detailed software and hardware architecture for a force feedback slider (FFS). To build applications incorporating a FFS, we offer developers a simple Application Programming Interface (API). In our realization, all time critical operations are implemented in hardware. An important goal of our work was to realize a generic and inexpensive FFS suitable for a wide range of applications. Such applications may serve as a complement to traditional graphical user interfaces (GUIs). One such application is presented, where the mechanics of a physical model is controlled and sensed through a FFS. Future challenges are partly related to performance issues, partly to the construction of user effective and inexpensive applications of FFS. Introduction Most computers require touch-based input combined with visual feedback. Optimal employment of visually presented information is a nontrivial task which gets harder with increasing system complexity. Considerable research effort has been invested into 3D information visualization. Even data with more than three dimensions may be reduced and presented in 3D. However, in models with a multitude of equally important factors, for instance from economics or biology, reduction to 3D will mostly loose out on complexity and system dynamics. This is one reasons that a combination of touchbased input with tactile or force feedback output has been offered much attention recently [1] [2] [3] [4] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] . Another reason is that the human hand can carry out complex tasks such as sensing and control with high performance and precision. The work presented in this paper one approach towards a generic architecture for the use of a force feedback slider (FFS). The work relies on hardware and software implementation and offers a simple Application Programming Interface (API) to enable integrated FFS applications. We have recognized that experts calibrating complex filters for RF applications perform their job effectively and fast. This motivated us to see the data manipulated in such tasks as coordinates of an appropriate vector space. With multiple sliders, we may offer such experts to work more directly with the multidimensional vector spaces. Additionally, grid level in these vector spaces may be sensed more directly, for instance through the use of senses slots. The most things we do in real life is connected to a feedback. It could be visual, audible or haptic. If we touch a door knob, we know what we have to do without to see what we do. That is an advantage we are not often using in computer work. Everything is going over the visual channel. So the visiual channel is overloaded, that could entail that the user do not recognize important information. But if you could use a force feedback device for what you are doing, the work will be safer, faster and more efficient. To test this we developed a prototype. In the next chapters we describe the hard and software and give results of our work. This paper first describes our model of the Force Feedback Slider , followed by a section on System design, a section results and discussion and future prospects. Figure 1: Force Feedback Slider , [7] Model of a Force Feedback Slider The control logic for a Force Feedback Slider has realtime requirements and thus it is especially important to have a carefully designed model. The API is expected to be on a high abstraction level for straight forward implementations. A system is required to translate this commands to the hardware in realtime. This section describes the operating modes, hardware and the transfer functions in slider driver. Operating Modes The following operating modes of a Force Feedback Slider , first suggested in [5], are: Position The fader is used only as input device, the motor is switched off. The user can move the fader as he likes. Elasticity A default position and maximal force is set. The user has to surmount a force. If he releases the handle, it goes back to the default position. Friction A default position and maximal force is set. The user has to surmount a force. But if he releases the handle it stays in that position. Gradual A number of discrete steps is set and the handle catches them. Texture A high frequency low intensity vibration is superimposed to the handle. This should give users an impression of a surface. Oscillation Emulation arbitrary movement. These operating modes are an abstract description of the capabilities of a Force Feedback Slider . As this is an abstract description they serve as an API. Hardware We use a standard (ALPS) motorized slider where we can just apply a force and detect the actual position of the handle. Most important task is to get the non-linear behavior of the motor and the friction under control. Position and force is a minimal but complete interface to the Force Feedback Slider . e.g. you could perform any operation the hardware is capable of. Transfer functions in Slider Driver The slider is driven by two inputs, these are position and force. The resulting output are position and force too. Depending on the operating mode, which will be applied, the configuration of inputs and outputs are different. As it is mention above, e.g. in the mode Position the output is only dependable on which position is adjusted on the slider. The force is set to zero. With two inputs and two outputs, there are theoretical 16 possible configurations. But it has shown that three configurations are enough, because of logical and practical reasons. The three transformation are the following: • output position1 to input position2 • output position to input position and input force 1the actual position of the slider or/and the preset default position from the system 2the new position value for the slider • output position and output force to input position and input force So it is possible to describe two orthogonal dimensions with one slider. (position and force) By using more then one slider, the number of dimension can be increased. The first dimension (position) is the linear movement of the slider and it is easy to see. The second dimension (force) can only be detected over the haptic channel of the finger. There is a simple limitation of increasing of the dimensions. Without training it is impossible to feel more then one or maybe two force fields. But with three or four dimension, which must not be detected over the visual channel, that could be an enormous simplification. To build an application with the slider, we defined a set of functions, which allows us to implement all operating modes discussed above. For this purpose we developed a data flow graph. On the left hand side in Figure 2 are the in and outputs of the application to the computer. On the other side respectively are the in and output to the Force Feedback Slider . Each mode uses a different set of functions (we call these transfer functions) and multiplexer settings. Transfer functions are: Function f1() is used to map the current position of the slider to a specific force. Every finite function is allowed. Function f2() takes the deviation of the current position and the application’s default position. This deviation is mapped to a definite force. Every force-field are allowed as long as the field is finite. Function f3() allows to take the current position and map it to a new default position. Table 1 lists the multiplexer settings and functions for every operating mode. Where P denotes a parameter and × is don’t care. With this model, we have a solid basis for further development. Design of the System Since the Force Feedback Slider has a high complexity, we decided to build the prototype in steps. Figure 3 presents the model with lower abstraction levels at the bottom, higher at the top. Each layer has a defined interface to the other. So it is possible to design each layer for itself. We decided to do the down up approach. We built the hardware [5] for the slider with a P-Regulator. (see section ) This section describes first the Physical Model, followed by the Layered Design, P-Regulator, Slider Driver and the API. Damped-Mass-Spring A Force Feedback Slider consists of a knob where the user can interact and feel the force. This knob is connected to a position sensor and a motor. There is also bearing which limits the degree of freedom to one. (only movement in one direction is allowed)

3 citations


Book ChapterDOI
01 Jan 2005
TL;DR: In this paper, the authors describe the development of a tangible user interface (TUI) called "Augmented Chemistry" (AC), giving details of the system's basic and specialized interactive tools, and outlines its educational context.
Abstract: This article reports on the many ways that tangible interaction can benefit chemistry education. It describes the development of a tangible user interface (TUI) called “Augmented Chemistry” (AC), gives details of the system’s basic and specialized interactive tools, and outlines its educational context. With these basic tools, users can choose elements from a booklet menu and construct 3D molecular models. The system can currently be used to teach certain aspects of organic molecular chemistry such as the octet rule. Expertise from optics, mathematics, molecular chemistry, software engineering and 3D programming played a role in the design of the system, making it a truly interdisciplinary project. We are currently planning to evaluate AC’s capacity to support effective, user-friendly chemistry education at a vocational school for trainees in the chemical industry. We will compare the system’s educational effectiveness with that of the institution’s established teaching methods.

2 citations