Smart art-sensor systems (DET.5100)

The research project concerns the development of new low-cost sensor systems, which enable human-computer interfacing for various art applications. The sensor systems will be applied in and for art performances as well as for art development. The field of art includes dance, painting, music and multi-media dance/sound/light performance.
The project will take advantage of the present activities of TU Delft in the field of capacitive sensors, and will be focussed towards the development of gesture sensors and personal position sensors. The gesture sensors make use of micro-machined capacitive accelerometers. A large number of these very small sensors are connected to body parts of art performers, for example dancers. After local signal processing the signals are wireless transmitted to a host computer. The personal detectors are based on the use of large electrode arrays which trans-capacitances are affected by the presence of humans. Such a technical system is intended for use in special prepared rooms, such as studios. The work will be organised in two sub-projects: a) smart accelerometer systems and b) smart capacitive position detection, which will be described now.

1. Smart accelerometer systems Small micro-machined smart sensors will be used to measure 3-dimensional acceleration. After integration and double-integration also speed and position information are found. To determine the integration constants, additional (reference) measurements are required. These additional measurements are performed at certain reference positions and moments, using special reference sensors. The whole sensor system should be fast enough to enable real-time use in art performances. The (sub-)project will include the following research items:
   
   • The overall system architecture in relation with the intended applications.
   • The selection, development or fabrication of low-cost accelerometers. When possible, existing elements will be used.
   • The selection, development or fabrication of low-cost reference sensors to measure starting position and speed. When possible, existing elements will be used.
   • The development of the hardware for low-cost signal processing circuits.
   • The development of dedicated software for the real-time smart data processing. The software will be implemented in microcontrollers, digital signal processors (DSP's) and personal computers.
2. Smart capacitive position detection
   In this sub-project the position of persons or body parts will be detected, using multiple-electrode capacitive-sensor systems [1]. The electrodes consist of low-cost conductive material, such as aluminium foil, which is connected to the walls, the floor and the ceiling. Between each pair of electrodes there exists an electric field which will be affected by the presence of objects or persons. With a multi-electrode capacitive sensor system [1] all of the trans-capacitances are measured. The acquired data is pre-processed in a microcontroller or DSP. The resulting data is transferred to a personal computer, which will do the final signal processing and output signals for light and sound controllers, and special actuators. The whole sensor system should be fast enough to enable real-time use in art performances. The (sub-)project will include the following research items:
   
   • The overall system architecture in relation with the intended applications.
   • The design of the electrode configurations, calculation and optimisation of the electricial fields and the effects of objects and persons for these fields.
   • The design of a fast electronic signal processor to modify, amplify, multiplex and digitise the electrode signals.
   • The design of data-processing methods and algorithms to derive detailed and selective information about position and movement of persons and objects from the trans-capacitance values.
   • The development of dedicated software for the real-time smart data processing. The software will be implemented in microcontrollers, digital signal processors (DSP's) and personal computers.

Meer informatie is te vinden op de website van de onderzoeksgroep.

Bij het project zijn vier bedrijven betrokken.