Projects on SPARC Lab

In-and-Out

Mon, 01 Jan 2024 00:00:00 +0000

The “In-and-Out” project aims to integrate indoor and outdoor spatial computing with cognitive research. Over the last decade indoor and outdoor spatial computing - previously developed in separation - has been integrated on the technical level. However, cognitive research has not followed this change. For example, wayfinding systems work indoors and outdoors but they do not change the type of support they provide based on the indoor/outdoor context and they do not consider outdoor information that is visible from the inside of the building. Thus, there is a gap between technological capabilities and cognitive understanding in spatial computing. This is critical because the key role of many spatial computing systems is to support cognition of their users, as in the cognitive geoengineering paradigm. This project plans to develop two prototypes: a wayfinding support system and an architectural design support system that for the first time will consider both indoor and outdoor information in an integrated manner. The interdisciplinary approach involves empirical experiments in VR environments and Bayesian statistical modelling.

SPACE-EYE

Wed, 01 Apr 2026 00:00:00 +0000

Current models of human–building interaction either analyse visual attention independently of movement, or reduce movement to a fixed field of view aligned with walking direction. This artificial separation leads to research results that are often trivial and therefore underused in practice. SPACE-EYE integrates theories of visual attention — visual saliency, the perception–action loop, embodied attention — with movement data, and develops a computational framework that models gaze and locomotion jointly.

What does it take to learn a building?

Mon, 09 Sep 2024 00:00:00 +0000

Large public buildings such as airports, hospitals, and shopping malls routinely prove difficult to navigate. Architects use analysis software to anticipate such issues, but existing tools operate at the scale of the whole building: they reason about layout legibility in aggregate, or run Agent-Based Models that oversimplify human cognition by, for example, keeping the field of view constant across all agents. Small differences in how an individual actually walks through a space are treated as noise, even though walking at a slightly different angle past a column might conceal a corridor opening behind it and cascade through the rest of that person’s experience in the building.

3D Mental Representations for Wayfinding

Fri, 13 Feb 2026 00:00:00 +0000

Wayfinding in complex 3D structures such as multi-level buildings, transport hubs, and shopping malls depends on mental representations that current research struggles to capture. Two communities work on this problem largely in parallel. Geoinformatics develops computational and Virtual Reality tools that probe spatial knowledge externally, typically through sketch mapping and related externalisation techniques. Architectural and psychological research, in turn, administers domain-specific psychometric tests to characterise spatial ability internally. The two traditions have rarely been brought into direct contact.

Visual Attention in Art Galleries

Mon, 01 Jan 2024 00:00:00 +0000

Art galleries are a very unique type of space: white, empty walls and the art itself are almost the only visual stimuli around. In such a situation, their spatial layout (the spatial arrangement of artworks and rooms) is what really steers the visitors’ attention.

The layout of space can have a larger impact on visitor’s attention than the differences between individual artworks. Although the layout can change the strategy with which visitors attend to artworks, cumulative time spent on attending to each picture is unaffected by its location. Methodologically, the project combined Mobile Eye-Tracking with the architectural theory of Space Syntax.

Architectural Cognition

Mon, 01 Jan 2024 00:00:00 +0000

How can psychology aid architecture in improving the usability of buildings? We study how space guides our behaviour, our attention, and our thinking. We design formal measures for architectural computation that are grounded in the cognitive experience of space.

In recent work, we have extended a popular architectural measure of visibility, called an “isovist”, from 2-D to 3-D environments. We demonstrated that our measure better predicts how “spacious” and “complex” a building feels to a human occupant, compared to the previously available methods.

3D Sketch Maps

Mon, 01 Jan 2024 00:00:00 +0000

The 3D Sketch Maps project, sponsored by the Swiss National Science Foundation within their Sinergia programme focusing on “interdisciplinary, collaborative and breakthrough research”, investigates 3D sketch maps from a theoretical, empirical, cognitive, as well as tool-related perspective, with a particular focus on Extended Reality (XR) technologies. Sketch mapping is an established research method in fields that study human spatial decision-making and information processing, such as navigation and wayfinding. Although space is naturally three-dimensional (3D), contemporary research has focused on assessing individuals’ spatial knowledge with two-dimensional (2D) sketches. For many domains though, such as aviation or the cognition of complex multilevel buildings, it is essential to study people’s 3D understanding of space, which is not possible with the current 2D methods.