[Back]


Talks and Poster Presentations (without Proceedings-Entry):

D. Grejner-Brzezinska, G. Retscher, C. Toth:
"Application of Artificial Intelligence and Multi-sensor Systems in Navigation and Engineering Geodesy";
Poster: IUGG 2011 Conference Earth on the Edge - Science for a Sustainable Planet, Melbourne; 2011-06-28 - 2011-07-07.



English abstract:
The ability to determine one´s position in absolute or map-referenced terms, relative to objects in the environment, and to navigate to a desired destination is an everyday necessity. Recent years brought up an explosion in the development of portable devices that support this functionality as well as provide a basis for the development of the Location Based Services (LBS). Most of the time, a Personal Navigation Device (PND) which combines the positioning and navigation capabilities, usually provided by the Global Positioning System (GPS), is used.
Pedestrian and personal navigation systems require continuous positioning and tracking of a mobile user with a certain positioning accuracy and reliability. However, navigating in urban and other GPS-impeded environments is a very challenging task. Thus, personal navigation systems require multiple navigation technologies to be integrated together, in order to serve as many different environments as possible for seamless and reliable navigation. Example technologies suitable for multisensor solutions supporting personal navigation include (aside from GPS/GNSS) ground-based RF systems, such as pseudolites suitable for confined and indoor environs, as well as cellular phone positioning for absolute position determination, dead reckoning sensors (e.g., gyroscopes, accelerometers, barometers or magnetic compass) to determine orientation, distance traveled and height. For location determination of a pedestrian in multi-storey buildings the Wireless Local Area Networks (WLAN), transponders or beacons installed in the buildings are increasingly used. Other indoor positioning systems include so-called Active Badge Systems (ABS). These methods can provide few-meter accuracy for indoor tracking and positioning. Robustness of the ultra wideband (UWB) signal to multipath fading and its high penetration capability makes it another technique suitable for indoor positioning. Other methods considered in indoor navigation are based on optical tracking systems, also referred to as image-based systems, and laser ranging systems, which provide range measurements to active or passive targets.
Recent years brought new developments in Artificial Intelligence (AI) techniques leading to an exponential increase in the number of applications in numerous areas, such as engineering, social and biomedical fields, as well as pedestrian navigation. AI is the study and design of intelligent agents, where an intelligent agent is a system that perceives its environment and takes actions which maximize its chances of success. In particular, AI techniques are very suitable in applications related to human motion modeling, and are being increasingly used for this purpose, due mainly to the complexity of the biological systems as well as the limitations of the existing quantitative techniques in modeling. Using AI methods allows for better process control and more reliable prediction/modeling of the processes under consideration. Examples of algorithms and methods used in AI are Artificial Neural Networks (ANNs) and Fuzzy Logic (FL). Both methods are suitable for modeling human locomotion dynamics that can be considered as a source of navigation information, where step length (SL) and step direction (SD) are the primary measurements that support navigation in a dead reckoning (DR) mode. The human dynamics model is calibrated while other sensors, primarily GPS, provide continuous navigation solution, and the human-based sensors are used in situation where other sensors cease to operate.
This paper provides a brief review of the navigation techniques suitable for pedestrian navigation, including some AI methods, with the focus on design and implementation of a Knowledge-Based Systems (KBS) for Dead-Reckoning (DR) navigation supported by the human locomotion model and a multi-sensor integration. In addition, applications of AI methods in engineering geodesy such as geodetic data analysis, deformation analysis, navigation, deformation network adjustment, optimization of complex measurement procedures and landslide monitoring using LiDAR data and ANN modeling, will be discussed.

Keywords:
Multi-sensor, Integration, Sensor fusion

Created from the Publication Database of the Vienna University of Technology.