En passant Coupon Collection

Spontaneous interaction in public places has evolved as crucial concern in interaction design, particularly in the domain of public advertising. Implicit interaction is a mode of spontaneous interaction in which the user is not attentively or explicitly expressing input to a system, but the system from observing the user, and considering all the available information that describes the user situation, is used to generate input and behaves accordingly. We have investigated implicit interaction as a potentially effective means to design for spontaneous interaction in the public. An en passant coupon collection system, CCS, has been developed to implement an implicit interaction based pervasive advertisement process. CCS is based on scanning and identifying Wi-Fi hotspots by their BSSID, and has been implemented for internet enabled smart phones. Scenarios demonstrate how the system collects bonus points, rewards, sales coupons, reward coins, etc. by users just walking by points of interest like stores, bus stops, schools or offices. CCS neither not being reliant to a preconfiguration of the Wi-Fi infrastructure, nor its use for wireless data communication, encourages for a parasitic use of existing Wi-Fi networks (already available in public places like shopping malls, airports, train stops, stations or city centers) as a pervasive advertisement system. 1 Implicit Interaction and Public Displays Pervasive Computing has developed a vision where the “computer” is no longer associated with the concept of a single device or a network of devices, but rather the entirety of situative services originating in a digital world, which are perceived through the physical world. It is expected that services with explicit user input and output will be replaced by a computing landscape sensing the physical world via a huge variety of sensors, and controlling it via a plethora of actuators. The nature and appearance of these services will change to be hidden ”in the fabric of everyday life”, invisibly networked, and omnipresent. They will greatly be based on the notions of context and knowledge, and will have to cope with highly dynamic environments and changing resources. Interaction with such computing landscapes will presumably be more implicit, at the periphery of human attention, rather than explicit, i.e. at the focus of attention: ”we will be able to create (mobile) devices that can see, hear and feel. Based on their perception, these devices will be able to act and react according to the situational context in which they are used” [Sch00]. Accordingly, as from the users point of view, ”implicit human computer interaction is an action performed by the user that is not primarily aimed to interact with a computerized system but which such a system understands as input.” [Sch00]. Implicit interaction is hence based on two main concepts: perception and interpretation. Perception is information gathering about the environment and situations, usually involving (technological) sensors. This information is generally provided implicitly to the system and displayed naturally to the user. Interpretation is the mechanism to understand the sensed data. Conceptually, perception and interpretation when combined are described as situational context. Clearly, pervasive advertisement, like e.g. advertisement via public displays, is a case demanding for principles of implicit interaction. Public display systems (PDS) [FV02], often referred to as Out-of-Home Media Systems, by gaining from technological advances in display technology, but also wireless communication technologies, have emerged as such a landscape, provisioning and controlling content related to context. ”Context” refers to any information describing the situation of an entity, like a person, a thing or a place, so the content being displayed in a PDS (in e.g. shopping malls, bus and train stations, airports, offices, schools, public places, etc.) can be adapted wrt. to all these types of information. While user interaction with PDS was not intended over their first generations of existence, the upcoming electronically operated and digitally controlled DSPs have encouraged a variety of solutions for explicit interaction [SD08], [BRS05], [BBRS06], [VB04] mostly being based on mobile phones as the interaction device [HR08], [WFGH99], [TCF07], the primary types of interaction being navigation and browsing [RYL07] [WK06]. Interaction with such PDS landscapes will presumably be more implicit, at the periphery of human attention, rather than explicit, i.e. at the focus of attention. With this paper therefore we address implicit interaction with PDSs, addressing the issues of a seamless, non-disrupting, yet situation aware modes of interaction in public advertisement. We have developed a Coupon Collection System (CCS) based on (even noncooperation) scanning for public Wi-Fi hotspots, and the mapping of unique hotspot IDs to geo-referenced advertisement services. The Coupon Collection System is developed in Section 2, and the roles of stakeholders engaging in a pervasive advertisement service system are related in Section 3. The corresponding software architecture and implementation details are outlined in Section 4. Section 5 exemplifies the use of the CCS system with smart phones as context aware CCS clients. Conclusions are drawn in Section 6. 2 The CCS Coupon Collection System CCS is divided into three major components. There are stationary components, which are installed on places where a user should be able to collect coins. The users carry the mobile components, which detect the stationary components automatically, without any explicit interaction. Finally, a server is needed, to manage the content which will be delivered to the mobile components. 2.1 Stationary Component WI-FI Access Points The major task of a stationary component is to be detected by the mobile component in a specific radius around its geographical location. One option to do that, without provoking an explicit interaction by the user, is to broadcast a radio signal with limited range Fig. 1. Of course, if there are two or more stationary components with overlapping radius, the interference between that signals has to be considered. To distinguish between different stationary components, each stationary component also needs a unique identifier. There is an existing and also very common technology which meets the requirements, the 802.11 wireless LAN access point. A functionality provided by access points is the beacon broadcast also known as SSID-Broadcast-, a specific package, sent at certain intervals. Usually the beacon broadcast is meant to make the access point known to wireless client devices in vicinity, therefore the beacon packages are not encrypted. Such a package contains, among other information, the MAC address of the access point, the name of the network, the time interval between the broadcasts and the encryption of the network [IEE07]. Using the beacon package, a client device is also able to evaluate the signal strength. Important for this system is mainly the MAC address of the access point for the use as a unique identifier. Usually, the interval between the beacons is set to 100 milliseconds. That leads to a very fast possible reaction time for the client. Other technologies are much slower, for example Bluetooth (with query scan interval set to default, the reaction can take up to 2.56 seconds [SIG09]). Also the frequency bands which are used by 802.11 Wi-Fi networks (2,4 and 5,18-5,7 GHz) are usable without a license. Figure 1: access points with SNR visibility: (from left to right) (i) one AP, (ii) multiple APs, (iii) multiple APs with overlapping SNR visibility, (iv) Possible Distribution of Wi-Fi Access Points in public space The interference management is provided by the 802.11 standard and supported by the hardware of access points [Ohr03]. That fact is important for the scalability of the system, because thereby overlapping domains are no longer a problem. The effective range of wireless LAN depends on factors as the transmission power, the used antenna and of course the consistency of the surroundings. As a reference value, 802.11b standard access points can reach a radius of up to 100 meters. To limit the radius of the access points it’s either possible to decrease the transmission power, which is does not work on every access point, or to filter signals under specific signal strength on the side of the mobile component [Ohr03]. If the signal strength of an access point is under certain value, the mobile component will just ignore it in first place or doesn’t redeem an BSSID for a coin when connected to the server. This feature doesn’t exist in the prototype by now, but it’s a likely target for further development. As an disadvantage, specification of a collection area for a specific virtual coin is not very accurate. It’s not possible to define precise boundaries which separate one collecting area from a non-collecting area or an area of another coin. For operating the CCS one has to be aware of this disadvantage and compensate it by clever distribution of access points and mapping of coins. In areas where there is already a wireless LAN infrastructure , the operator can call on existing access points. In fact, every access point with enabled beacon broadcast can be used to act as a stationary component in this system, even if it doesn’t belong to the operator of the system. 2.2 Mobile Component Smart Phone The requirements a mobile component has to meet to work in this system are matched by many of the current and future generation smart phones. They provide the required connectivity like wireless LAN for collecting BSSIDs and UMTS for connection to a server on the web, more than enough computing power and storage, as well as high resolution displays and keys or even touch-screens for interaction with the software. However, not every smart phone is equally well qualified for the use in this system, even if it meets all hardware requirements. The mobile component of the system is implemented as software on a smartphone. If the user wants to start collecting, he has to activate the software. From that point of time, until he switches it off again, the software collects coins. An important point which has to be considered is the increased power consumption. Without a program running, a smart phones switch into an energy saving mode, where all but the most basic functions are deactivated. While the software searches for access points, the phone will not be able to do that, which leads to a significantly reduced operation time. The prototype for the mobile component of the system was developed for the Apple iPhone, a quite innovative and widespread piece of technology.