Riders Have Spoken: Designing and Evaluating an Archive for Replaying Interactive Performances

Lead Research Organisation: University of Sheffield
Department Name: Information School


In recent years, a number of organizations within the field of electronic arts have devoted attention to the task of archiving and preserving, often transitory, electronic works of art and digital performances. These organizations have tried to address some of the problematic aspects of archiving and preserving electronic media, e.g. the research and development process that is often involved, user interactions, distributed authorship of the work, and dependency on hardware and software components, by developing documentation strategies and metadata models that aid in accessing the materials and increasing their interoperability. In addition interactive games and performances present particular challenges for capturing, archiving, and replaying. These challenges relate to the often distributed nature of the artwork, its open design, the multiple participants involved, and the heterogeneous nature of the data, e.g. audio files, video files, GPS data, generated by the performance. Trying to capture the live character of such performances is extremely difficult. Rider Spoke, the mixed reality interactive performance that we intend to archive and replay, was developed by the artists Blast Theory in collaboration with the Mixed Reality Laboratory at the University of Nottingham as part of the European research project Integrated Project of Pervasive Games (IPerg). We chose this particular work because it exemplifies the challenges posed by the documentation of interactive performance. A Rider Spoke performance takes place in both virtual and real spaces, and uses mobile technologies and locative media in dramaturgically complex ways. Since its launch in 2007 the work has so far been experienced by over 2000 participants in London, Athens, Brighton, Budapest; and most recently Sydney and Adelaide in 2009. Designed specifically for cyclists, Rider Spoke combines elements of performance, gameplay, and interactive technology by inviting participants to cycle through the streets of a city equipped with a handheld computer that is mounted to their bicycle. Like other performances, Rider Spoke contains a fixed set of parameters and rules within which game action takes place; unlike the scripted performance of a play, however, Rider Spoke is constituted as much by the emergent social and interpersonal interactions between 100 or more so players, as it is by players' following pre-specified game rules. Hence in the course of a single performance multiple players will use mobile and location-based technologies to mediate their interactions and in doing so generate different types of data that need to be captured in the archive, along with any materials relating to the research and development process that was used in the creation of the work, and any materials relating to the reception of the work e.g. reviews. In developing an archive from a heterogeneous set of materials, it is crucial that attention be given to developing a consistently assigned set of descriptors, i.e. a metadata scheme, in order to enable subsequent users of the archive to access the materials in a meaningful and orderly way. The main outcomes of the research will be: (i) an understanding of the problems and issues related to the capturing, archiving and replaying of interactive performances, specifically Rider Spoke (ii) a set of requirements for archiving and replaying Rider Spoke; collectively developed by the project team, along with contributions from archivists, and from humanities researchers. (iii) The construction of a metadata schema and vocabulary of terms for the consistent description of the materials in the archive (iv) a functional operational prototype that enables users to navigate within and cross each layer of the archive (v) outcomes and lessons learnt from the process of archiving and replaying Rider Spoke, and how this informs the capturing, archiving, and where appropriate replaying of electronic artworks and interactive performances in general.


10 25 50

publication icon
Foster, J. The Relational Archive in Duration (Before and) After Media

publication icon
Foster, J.; Price, D; Benford, S.; Giannachi, G. Navigating beyond text: On conversation and visualization in the development of a digital archive in Digital Resources in the Humanities and Arts

Description The archiving of electronic artworks and interactive performances presents considerable challenges of capture and documentation, information organization, and interaction. The RHS project brought together an interdisciplinary team from information science, human-computer interaction, performance studies and computer science to explore these challenges in the context of archiving a work called Rider Spoke by British Artists Blast Theory (http://www.blasttheory.co.uk/bt/work_rider_spoke.html). The interdisciplinary nature of the team enabled different kinds of disciplinary knowledge to be brought to bear on the development of the archive e.g. information organization and subject access, domain knowledge of the content of the work, media replay and interface design.
Exploitation Route The specific work around which a number of different prototype digital archives was developed was an interactive performance work by British Artists Blast Theory Called Rider Spoke. This can be taken forward by considering a) the participation of the public in the performance of other similar works b) the capturing of these events and how they are represented in different versions c) the organization of and access to the resulting digital archives d) their networked distribution e) the possibility of either creating new pathways through the existing archive or creating new trajectories (narratives).
Sectors Creative Economy,Digital/Communication/Information Technologies (including Software),Education

Description A distinctive contribution was made through to addressing these challenges through the use of an iterative, prototyping approach that enabled the incremental design and evaluation of a usable archive. The development of the archive, based on a work called Rider Spoke by British artists Blast Theory (http://www.blasttheory.co.uk/ bt/work_rider_spoke.html) underwent two major iterations during the course of the project: an initial Cloudpad prototype design and evaluation at San Francisco Art Institute, Stanford Libraries, and San Jose State University School of Library and Information Science; and a further Cloudpad prototype design and evaluation at the Information School, University of Sheffield. The initial Cloudpad prototype was developed completely from scratch and designed as a classic web application with a Model-View-Controller architecture using the Spring Framework (see http://www.springsource.org/). The prototype was modelled on the concept of a temporal trajectory, that is, the system allows documents to be authored that follow a linear timeline punctuated by events. CloudPad documents are referred to as trajectories. Trajectories are built up from a palette of events; display/playback of audio/video elements and display of text annotations. This allows text annotations to be synchronized with media playback as well as for multiple media streams to be synchronized with one another. Media and annotation elements can be positioned relatively to one another on the screen. Figure 1 shows two video elements being synchronized with one another and an annotation being authored. The materials in the prototype consisted of raw media content that included interviews and video documentaries, or trajectories, shadowing the journeys of different riders. These trajectories were subsequently annotated by the artists. The initial San Francisco Bay Area evaluation included feedback from 16 students and staff at a number of institutions in the San Francisco Bay area, on the interface and the different experiences they had watching pre-made trajectories vs. creating their own trajectories. These user experiences led to a further development of the prototype and its evaluation at the Information School, University of Sheffield. Besides the common interface improvements and bug fixes two new substantial additions were made: search facilities and a structure editor. A search facility was developed that allowed for a user to search annotations in trajectories for words or phrases. Trajectories that contained annotations containing the word or phrase were returned by the search. In addition, search terms were recorded and displayed in a tag cloud to visualize the most common search terms. Search terms could also be highlighted when replaying trajectories. The structure editor provides an alternative view on a trajectory. The user selects a trajectory to view in the structure editor, they are then presented with a list view of the annotations in that trajectory in time order. It is then possible to reorder the annotations, add new annotations and to add two different types of structure separators. The first is the question separator that can be used to mark the points at which the rider in the trajectory is listening to different questions in Rider Spoke. The second is the transition separator which can be used to mark when a transition between events in Rider Spoke occurs. Edited structures can be saved and viewed later. The materials for the Sheffield deployment consisted of the same materials for the Stanford Bay Area evaluation. 74 students were first invited to tag and annotate the video documentaries or trajectories; their tags and annotations were then analysed with a view to compiling a classification scheme and thesaurus that would enable consistent but flexible subject access to the content of the archive. Underpinned again by the concept of trajectory, analysis of tags and annotations led to the identification of 7 facets. The tags for each facet were then further categorized and where appropriate sub-categorized into hierarchies and sub-hierarchies. There are currently 24 hierarchies, 22 of which were developed from a thematic analysis of the accumulated tags with 2 (disciplines, events) taken from the Getty Vocabularies' Art and Architecture Thesaurus. Incorporation of the tags in the scheme enables pointing to and precise recall of a particular data point within one of the trajectories in the archive. At the same time each hierarchy acts as a metadata resource enabling retrieval of one or more topic-related data points within or across trajectories (e.g. all responses to Rider Spoke question 'Tell me about your father'). Hierarchies or facets can also be browsed if a user has an interest in the topic of a particular hierarchy (e.g. all actions or interactions) or facet (e.g. aesthetics) (see Figure). The schema can also be used as the basis for faceted search that combines hierarchies from different facets (e.g. participants + mood). The classification scheme can also be used flexibly to act either as a controlled vocabulary for the archival assignment of metadata pertinent to the trajectories; or as a way of grouping user-driven tags that might be represented as a tag cloud, into meaningful categories. The final prototype archive will be released in two formats, the source code will be hosted on Github (https://github.com/dominicjprice/Cloudpad) and a precompiled archive will be made available as an Amazon Machine Image.
First Year Of Impact 2010
Sector Creative Economy,Digital/Communication/Information Technologies (including Software),Education
Impact Types Cultural

Title Cloudpad 
Description Cloudpad is a web application for the archiving of 'experiences', for example Blast Theory's Riderspoke. It allows media recording during live experiences to be imported into the system and then played back in a non-linear fashion. Media objects can be set to play at specific times in the document timeline, synchronised with each other and text annotations added to explain the 'experience' of the document editor. It allows for new experiences to be created from the recordings of other experiences. 
Type Of Technology Software 
Year Produced 2011 
Open Source License? Yes  
Impact Cloudpad is mainly targeted at people who wish to archive performances that have no central object, for example in a play the script is the central object. Contributed to the development of expertise in archiving experiences at the Mixed Reality Lab, University of Nottingham as part of its Digital Economy work. 
URL https://github.com/dominicjprice/Cloudpad