There is a new wave of pervasive computing applications, which synergistically fuse the sensing and analytics capabilities of personal wearable/mobile devices and cheap IoT platforms to capture fine-grained human and environmental context. In this talk, I will first describe and review recent/ongoing work on a diverse set of such wearable and IoT-assisted applications, including smartwatch-based diet monitoring, Bluetooth beacon-based in-store shopper behavior analytics, VR smartglass-based empathetic accessibility testing of mobile applications and wearable-based remote monitoring of factory machinery. Through these applications, I shall distill some of their common characteristics that distinguish them from conventional desktop/mobile “productivity” applications, such as: (i) the use of multi-sensor processing pipelines & supervised machine learning, (ii) real-time requirements to support interactive and immersive interfaces, and (iii) distributed cross-device execution of application components. The combination of physical sensing, wireless communication and computationally intensive processing in such applications leads to novel forms of application failure and non-performance. I shall conclude with some early thoughts on new testing capabilities and paradigms needed to tackle such wearable/IoT based cyber-physical systems and applications.