#	Pagina
attuale pagina	/open-h2020/projects/197223/results.html

Report

Teaser, summary, work performed and final results

Periodic Reporting for period 2 - WayTO (Wayfinding Through Orientation)

Teaser

Summary

Wayfinding is a task that we manage every day while going to work, visiting friends, or going on vacation. With the emergence of car and pedestrian navigation systems, we gained support for wayfinding tasks in unfamiliar environments. Although wayfinding research has gone through tremendous development, with cognitive aspects particularly attracting interest in research, it suffers from some fundamental shortcomings: State-of-the art wayfinding research still adheres to the principles of turn-by-turn navigation. At each decision point, users receive the next turn instruction guiding them towards the destination. The user has no orientation and no overview of the surrounding environment. Turn-by-turn navigation is cognitively not adequate: When humans do wayfinding, they do not execute instructions separately one after another, but they learn the spatial configuration during wayfinding and build up cognitive maps to orient themselves. Because turn-by-turn navigation solely communicates directions at decision points, it supports only the acquisition of route knowledge. It does not support the userâ€™s very own habit in gaining orientation in unfamiliar environment.

This project aims at developing new means of wayfinding based on orientation: Recent research enhanced navigation systems by making navigation instructions easier to understand through intuitive landmark information, through simpler decision points, and through easier routes. However, this research does not tackle the fundamental problems of turn-by-turn navigation. We argue that orientation information is more suitable for wayfinding assistance: It supports users in acquiring survey knowledge, because it communicates wayfinding information that helps users to mentally build up a two-dimensional map of her environment. This cognitive map is used for orientation. Being oriented on your way is a precondition to enable people to verify wayfinding instructions. A cognitive map of the environment allows users to adapt their way according to unforeseen changes, finding shortcuts, circumnavigating obstacles or spontaneously making detour trips to points of interest. Fig. 1 contrasts spatial knowledge learned from turn-by-turn wayfinding with spatial knowledge learned from orientation wayfinding. From turn-by-turn wayfinding, users only learn the sequence of turns that need to be taken. The survey knowledge learned with orientation wayfinding will allow the user to get an overview of the route and the surrounding environment and integrate various types of spatial information into her 2-dimensional cognitive map.

In this project, we have four objectives:

Objective 1: Scientific Understanding of Orientation Wayfinding. Research needs find out what defines orientation in humans and what kind of information supports orientation. We need to prove the hypothesis that cognitive maps and survey knowledge directly relate to orientation.

Objective 2: Automatic Generation of Orientation Information. For computer-supported wayfinding assistance we must be able to handle orientation information automatically. However, orientation information has some fundamentally different characteristics than conventional spatial data stored in geographic information systems. Orientation information does not have a consistent level of generalization, it is highly schematized and it refers to vernacular, vague places not included in traditional maps. Furthermore, the level of generalization and schematization of orientation information depends on the route.

Objective 3: Communication of Orientation Information. We will integrate orientation instructions into route directions. New visualizations will be required to account for the characteristics of orientation information: How can we schematize maps that visualize the route on the one side (route schematization methods) and survey information necessary for orientation (network schematization methods) on the other hand? How can we deal with inhomogen

Work performed

\"Major Achievements:

The project developed and published a detailed definition of orientation \"\"A situated, goal-driven interaction of the user\'s partial and uncertain knowledge, reasoning about spatial properties of the relevant subset of the environment, and the external information available to support it.â€ [Pub1] Furthermore, the extended the well-acknowledged taxonomy of wayfinding tasks by Wiener et al by adding to it and describing a cognitive task of \'oriented path following\'. We argue that such a cognitive task is one of the most often employed tasks by human navigators in real life, but rarely if ever studied in controlled experiments. In analogy to the taxonomy for cognitive wayfinding tasks we proposed a new taxonomy on wayfinding assistance systems including systems that are based on orientation wayfinding [Pub2].

We increased the awareness for orientation wayfinding in the scientific community by editing a special issue and organizing a workshop [Pub14, Pub15].

We have completed five experiments and several other studies are ongoing. With the empirical data that we collected, we could show
(i) that people extensively use orientation information in their wayfinding instructions [Pub3, Pub4, Pub13, Pub17].
(ii) orientation information has an effect on spatial knowledge of people and people structure (and distort) their cognitive map according to global landmarks and larger structures (orientation information). We could show that people integrate route information with survey information into a single map, i.e. a navigation system should work in a similar way embedding wayfinding instructions for a particular route into important (orientation) information about the surroundings [Sub1].
(iii) people are able to use our prototype for the future orientation wayfinding assistance system [Pub2].
(iv) An algorithm was developed to extract automatically orientation information from maps. The route dependence is still subject to further investigations. Preliminary results have been published at the AGILE conference (to be presented in June 18) [Pub5, Pub6, Pub7].

The results have led to 19 scientific publications listed below.
We also presented the project at major events and the local media has been interested in WayTO and published newspaper articles about the project.

Results and Achievements per Work Package and Task:

Work Package 1 (Staff: Schwering, Krukar, LÃ¶wen)
Task 1.1: We have been developing a work in progress definition on orientation and suggested a new classification scheme for wayfinding tasks and wayfinding assistance systems. Results have been published in [Pub1], [Pub2]. We extended the taxonomy of non-aided wayfinding tasks to aided wayfinding tasks, providing a novel theoretical perspective on the problem of using navigation support systems for different subsets of wayfinding activities. [Pub2].
Task 1.2: A study revealed that people extensively use orientation information in their study. People judged orientation instructions as helpful in wayfinding. The results of this study have been published in [P 3]. We conducted a comparison of spontaneous knowledge recall for classical turn-by-turn instructions versus orientation instructions versus skeletal instructions, developing a classification scheme for sketch maps [Sub1].
Task 1.3 and Task 1.4: Although LÃ¶wen started his PhD with a delay, he has already supported the project during his MSc degree as student assistant. In his MSc thesis, he has been working on a topic related to task 1.3 and 1.4. Results lead to [Pub11]. LÃ¶wen has now refined our originally defined classification scheme [Pub3] in the context of automatic generation of orientation information and usage of open data such as OpenStreetMap. This work lead to publications [Pub5, Pub6].
LÃ¶wen has submitted his PhD proposal, describing mainly research in work package 1 and 2, at the doctoral colloquium at 13th International Conference on Spatial Information Theory (COSIT) [Pub7].

\"

Final results

Existing research in cognitive wayfinding improved navigation in various respects, but never questioned the paradigm of wayfinding itself. We propose the new paradigm Wayfinding Through Orientation, where the navigation system supports the user in orienting, spatial learning and cognitive mapping. Although researchers agree that survey knowledge is equally important to route knowledge and that survey knowledge substantially contributes to cognitive mapping and orientation, so far (to the best of our knowledge) no researcher used survey or orientation information to improve wayfinding.
Even though orientation wayfinding has never been investigated by other researchers before, we believe that it will revolutionize the way of navigation: Wayfinding Through Orientation has a completely new understanding of a personâ€™s role in navigation. Orientation wayfinding involves usersâ€™ in the navigation process through addressing the usersâ€™ cognitive learning and thinking abilities. It makes users less dependent on the navigation system, more self-confident and sensible to their environment. Users can make informed choices and are able to adapt their route according to unforeseen changes. Users will be able to find shortcuts, circumnavigate new obstacles or spontaneously make detour trips. Being oriented and having an overview of the environment is necessary to verify and understand wayfinding instructions. In contrast, users of turn-by-turn systems solely execute given turn instructions.
Our research has profound impact in all the three disciplines involved: The new definition of orientation in wayfinding and the new methods to measure orientation contribute to cognitive science. The new approaches to capture, process, and communicate orientation information are major contributions for geographic information science and computer science. Our research lays the scientific foundations for future navigation systems following the Wayfinding Through Orientation paradigm.