We present novel accelerometer-based techniques for accurate and fine-grained detection of transportation modes on smartphones. The primary contributions of our work are an improved algorithm for estimating the gravity component of accelerometer measurements, a novel set of accelerometer features that are able to capture key characteristics of vehicular movement patterns, and a hierarchical decomposition of the detection task. We evaluate our approach using over 150 hours of transportation data, which has been collected from 4 different countries and 16 individuals. Results of the evaluation demonstrate that our approach is able to improve transportation mode detection by over 20% compared to current accelerometer-based systems, while at the same time improving generalization and robustness of the detection. The main performance improvements are obtained for motorised transportation modalities, which currently represent the main challenge for smartphone-based transportation mode detection.
Linear acceleration is an important enabler for many applications of mobile and wearable activity recognition. The most common approach for estimating linear acceleration is to estimate the gravity component of accelerometer measurements and to project gravity-eliminated accelerometer measurements onto horizontal and vertical planes. Consequently, the accuracy of the linear acceleration estimates is highly dependent on the accuracy and robustness of the underlying gravity estimation algorithm. The present paper contributes by developing a novel approach for gravity and linear acceleration estimation from accelerometer and gyroscope measurements. Our approach improves on previous solutions by (i) providing increased robustness in the presence of sustained acceleration; (ii) detecting and filtering out common types of noise, such as centripetal forces and shifts in device orientation caused by spontaneous user interactions; (iii) operating on shorter time windows, making our approach suitable for applications that require rapid updates on user activities; and by (iv) distinguishing between lateral and longitudinal components of linear acceleration. Experiments carried out using over 100 hours of measurements demonstrate that our approach results in significant improvements in the accuracy of linear acceleration estimates, and improves robustness against common sources of noise in the estimation process. Specifically, our method achieves over 40% improvements in the accuracy of reconstructing speed information and over 70% improvements in the accuracy of estimating travel distances.
Pervasive availability of programmable smart devices is giving rise to sensing and computing scenarios that involve collaboration between multiple devices. Maximizing the benefits of collaboration requires careful selection of devices with whom to collaborate as otherwise collaboration may be interrupted prematurely or be sub-optimal for the characteristics of the task at hand. Existing research on collaborative scenarios has mostly focused on providing mechanisms that can establish and harness collaboration, without considering how to maximally benefit from it. In this paper, we contribute by developing COSINE as a novel approach for selecting collaborators in multidevice computing scenarios. COSINE identifies and recommends collaborators based on a novel information theoretic measure based on Markov trajectory entropy. Rigorous experimental benchmarks carried out using a large-scale dataset of deviceto-device encounters demonstrate that COSINE can significantly improve collaboration benefits compared to current state-of-theart solutions, increasing expected duration of collaboration and reducing variability of collaborations.
With the advances in smartphone technologies, sustainable mobility has become an active research topic in the field of ubiquitous computing. We present a persuasive mobile application that automatically tracks the transportation modes and CO2 emissions of the trips of the user and utilizes this information to present a set of actionable mobility challenges to the user. A longitudinal pilot experiment with the system showed that subjects perceived the concept of challenges as positive, with constructive findings to inform further development of the application especially related to personalized challenges.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.