RELATE: Assessment of relative positioning technologies for compositional tangible interfaces

RELATE is a project funded by the European Commission in the Disappearing Computer Call (IST-2001-39264). Start date: Jan 2003

INTRODUCTION
To provide interaction with computing in ways that are more tangible, easier to grasp (literally and metaphorically), and better integrated with everyday environments is a key element in visions of more inclusive information technologies, e.g. ubiquitous computing and ambient intelligence. It is also an explicit objective in work programme 2002, and continues to be a research priority in FP6 toward creation of intelligent interfaces and surfaces that support seamless interaction between humans and devices, virtual and physical objects. Tangible user interfaces have recently attracted considerable research interest, e.g. with inspiring early work by Ishii (MIT Media Lab) and interactive artefacts explored in various FET-i3 and FET-DC projects. However, there has been no breakthrough yet from individual tangible objects to compositional tangible interfaces that support richer interactions, in particular the exploitation of spatial arrangement of objects as meaningful input to an interactive system. Two pieces of technology are missing: (i) ad hoc communication between tangible objects - this is now becoming available, e.g. with low-cost RFM, Bluetooth and platforms such as Smart-Its; (ii) ad hoc relative positioning of tangible objects at close range - this has not been addressed previously and is therefore taken up as research challenge in the proposed assessment project.
The problem that we seek to address is how to establish the spatial relationships between tangible objects that together form an interface. In the virtual realm, it is easy for an interactive system to maintain correspondence between the spatial relationships of objects on the screen and the underlying semantic relationships. However in the physical realm, it is a serious challenge to establish the spatial arrangement of objects. Only if this is solved in a scalable and widely deployable way we will be able to construct everyday interfaces in which semantic relationships can be manipulated through spatial arrangement. The current solutions fall short of easy deployment. They are either very constrained (e.g. requiring physical connection of objects rather than free placement to relate them) or dependent on intelligent infrastructure, e.g. vision systems to track objects, location sensor networks, or embedded ID tag readers. In contrast we seek to investigate technologies for peer-to-peer relative positioning among the objects themselves without any infrastructure requirement. It has been suggested that relative positioning will be available as by-product in ad hoc wireless networks, however it is important to understand that this does not work over very short distances. Yet, multiple objects that people interact with will typically be at very close range, therefore requiring investigation and development of a novel technology concept.

SUMMARY

• Design technology for detecting the relative position and orientation of small interface objects
• for the use on common interaction surfaces, e.g. tables
• spontaneous set-up of the user interface without the need for infrastructure, calibration and maintenance to enable novel compositional tangible interfaces
• Implement the technology on small devices (max 7x7 cm)
• Build a demonstration environment with at least 10 devices
• Enabling new ways of interaction
• Show these new ways of interaction by design applications that utilize the technology e.g. games, group-work support
• Implement the applications on the devices an test

OBJECTIVES
Our general aim is to develop relative positioning technology for compositional tangible interfaces. The envisioned technology is decentralized and embedded in the tangible interface components, not requiring any additional infrastructure and therefore working in arbitrary environments. Each tangible interface component will be augmented with the ability to measure its own orientation and its distance from neighbouring interface components. By means of communication, the distributed individual observations will be combined to establish the overall spatial layout of the tangible interface. In a first step toward such a technology we will consider spatial layout on surfaces, i.e. on 2D-planes rather than in 3D-space

As our objective is to develop a concept that is practical for deployment in everyday tangible interfaces, our focus is not on functional performance alone (accuracy of the relative position measurement) but on performance in relation to a variety of factors that need to be assessed:

• Speed: the utility and application space of a relative positioning technology depends on how fast the overall physical configuration of all user interface components can be acquired.
• Scalability: in what way does the number of interface components influence the acquisition speed and robustness of this technique? What is the upper limit of interface components for a comfortable real-time interaction?
• Robustness: what environmental factors (light conditions, magnetic fields generated in high tech environments, etc.) can have a negative influence on the performance of the system?
• Cost of implementation alternatives: for deployment in everyday interfaces it becomes important to achieve a cost effective implementation, and it will be important to understand cost vs. gain trade-offs.
• Applicability: the technology is to support the widest possible application range, and it is important to explore the application space and to assess in which ways technology factors may influence applicability.
  

TECHNICAL INNOVATION

Ad hoc relative positioning for compositional tangible interfaces has not been addressed previously. The anticipated system will be distinguished by the following set of characteristics:

• measure two technical parameters between objects/interface components: angle and distance
• networking of components to share observations
• facilitate use of spatial layout of tangible interface for meaningful user interaction
• work anywhere without infrastructure based on peer-to-peer communication
• support ad hoc integration of components to a tangible interface
• support practical deployment: low energy consumption for long run per battery/charge, small size/form-factor (max. 7x7x1.5 cm including battery), untethered handling (wireless communication), robustness in everyday interaction
• invisibility of the technology: no technical introduction needed, zero administration and readily usable in unaccustomed environment

The system will be implemented in small device modules. These can be used for construction of artificial interface objects, e.g. smart counters with versatile use, as well as for augmentation of existing objects, e.g. tools on a workbench, or magnets on a whiteboard. For different applications, device modules may be packaged differently into interface objects, with varying form-factors.

TEAM
RELATE is a joint project between Hans-Werner Gellersen's group at Lancaster University and TecO (University of Karlsruhe)