Public displays and mobile phones are ubiquitous technologies that are already weaving themselves into the everyday life of urban citizens. The combination of the two enables new and novel possibilities, such as interaction with displays that are not physically accessible, extending screen real estate for mobile phones or transferring user content to and from public displays. However, current usability evaluations of prototype systems have explored only a small part of this design space, as usage of such systems is deeply embedded in and dependent on social and everyday context. In order to investigate issues surrounding appropriation and real use in social context field studies are necessary. In this paper we present our experiences with field deployments in a continuum between exploratory prototypes and technology probes. We present benefits and drawbacks of different evaluation methods, and provide a number of validated lessons from our deployments.
TopThere is surprisingly little published work relating to the combination of mobile phones and situated public displays, and the vast majority of these systems have only been evaluated in the lab, if at all. ContentCascade (Himanshu, Gossweiler, & Milojicic, 2004) for example enables a user to download content from a public display onto her mobile phone using Bluetooth. The system was tested in a small and informal user study using movie clips. More recent work by Maunder, Marsden and Harper (2007) has investigated the potential for supporting mobile phone interaction with public displays in order to enable users to select and download content without requiring the user to keep their phone in a discoverable state. Their approach required the user to take a picture of the content screen that he/she wishes to download and then send this picture back to the public display server as a Bluetooth transfer, thus providing the server with the user’s phone’s Bluetooth MAC address. The server then performs image recognition to determine the content required by the user, which is then transferred via Bluetooth to the user’s phone. The system has only been evaluated informally. Ballagas, Rohs, Sheridan and Borchers (2005) present a survey of interaction techniques with mobile phones, most of which are used to generate input to a public display. The majority of systems they present have been evaluated only in lab studies. Rukzio, Boll, Leichtenstern and Schmidt (2007) present a comparison of different interaction techniques with mobile phones, which have been evaluated in the lab. Some systems use Bluetooth as a means to detect the presence of people rather than as a means to enable explicit interaction. Two examples of these systems are the BluScreen system (Payne, David, Jennings, & Sharifi, 2006), which links advertisement displays, agents bidding for advertisement space and the detection of presence via Bluetooth, and CityWare (Kindberg & Jones, 2007), where urban activities of users were tracked with Bluetooth scanners.
The majority of systems built have been evaluated only in lab settings and not in field deployments. Therefore, there is little knowledge to date regarding appropriation into everyday life of systems that combine public displays and mobile phones. This general bias has also been identified generally in mobile HCI research (Kjeldskov & Graham, 2003). The focussing on usability issues only and ignoring appropriation has come under increasing critique (Greenberg & Buxton, 2008). Although the added value of field studies for finding usability flaws can be doubted (Kieldskov, Skov, Als, & Høegh, 2004), its unique applicability to study appropriation has been shown (Rogers et al., 2007). For the related field of Ambient Displays, Skog (2006) has shown that many interesting aspects can only be observed from longitudinal field studies. For public displays, Huang, Mynatt and Trimble (2007) have shown that the challenges of the real world lead to unexpected usage patterns that probably cannot be predicted from lab experiments.