oDect: an RFID-based Object Detection API to support applications development on Mobile Devices

The RFID technology is becoming ever more popular in the development of ubiquitous computing applications. A full exploitation of the RFID potential requires the study and implementation of human\u2013computer interaction (HCI) modalities to be able to support wide usability by the target audience. This implies the need for programming methodologies specifically dedicated to support the easy and efficient prototyping of applications to have feedback from early tests with users. On the basis of our field-working experience, we have designed oDect, a high-level language and platform-independent application programming interface (API), ad hoc designed to meet the needs of typical applications for mobile devices (smart phones and PDAs). oDect aims at allowing application developers to create their prototypes focusing on the needs of the final users, without having to care about the low-level software that interacts with the RFID hardware. Further, in an end-user developing (EUD) approach, oDect provides specific support for the application end-user herself to cope with typical problems of RFID applications in detecting objects. We describe in detail the features of the API and discuss the findings of a test with four programmers, where we analyse and evaluate the use of the API in four sample applications. We also present results of an end-user test, which investigated strengths and weaknesses of the territorial agenda (TA) concept. The TA is an RFID-based citizen guide that aids\u2014through time- and location-based reminders\u2014users in their daily activities in a city. The TA directly exploits EUD features of oDect, in particular concerning the possibility of linking detected objects with custom actions