Abstract
Non-intrusive monitoring of fine-grained activities of daily living (ADL) enables various smart healthcare applications. For example, ADL pattern analysis for older adults at risk can be used to assess their loss of safety or independence. Prior work in the area of ADL recognition has focused on coarse-grained ADL recognition at the activity level (e.g., cooking, cleaning, sleeping), and/or course-grained (hourly or minutely) activity duration segmentation. It also typically relies on a high-density deployment of a variety of sensors. Finer-grained (sub-second-level and event-based) ADL recognition, could provide more detailed ADL information, which is crucial for enabling the assessment of patients' activity patterns and potential changes in behavior. To achieve this fine-grained ADL monitoring, we present a heterogeneous multi-modal cyber-physical system, where we use 1) distributed vibration sensors to capture the action-induced structural vibrations and their spatial characteristics for information aggregation, and 2) single point electrical sensor to capture appliance usage with high temporal resolution. To evaluate our system, we conducted real-world experiments with multiple human subjects to demonstrate the complementary information from these two sensing modalities. Our system achieved an average 90% accuracy in recognizing activities, which is up to 2.6x higher than baseline systems considering each state-of-the-art sensing modality separately.
Highlights
The Internet of Things (IoT) and its rapid development enables various smart home applications that have the potential to support independent living for older adults (Azimi et al, 2017; Kokku, 2017)
It is challenging to achieve fine-grained, which we define as sub-second-level and event/action-level, activities of daily living (ADL) recognition nonintrusively and sparsely because each ADL consists of several events or actions
Our ensemble approach achieved the highest classification accuracy for half of the events, and the average accuracy over eight events is 90%, which is a 1.5× to 2.6× improvement compared to the baselines (56, 35, and 61%)
Summary
The Internet of Things (IoT) and its rapid development enables various smart home applications that have the potential to support independent living for older adults (Azimi et al, 2017; Kokku, 2017). Activity type recognition methods leverage learning algorithms to improve the accuracy and robustness for classifying given sensing signals (Castanedo, 2013; De-La-Hoz-Franco et al, 2018) Combined, these efforts focus on context-level information with the timeresolution of minute or hour, which is coarse-grained. Prior attempts for fine-grained ADL monitoring combine electrical sensors and passive RFID sensors, where the on-wrist RFID provides locations and an electrical sensor provides appliance usage information. These methods (e.g., Fortin-Simard et al, 2014) require high-density sensor deployment and people carrying devices or tags during their activities. Especially those with cognitive impairments, may find it difficult to remember to wear or uncomfortable to wear such devices
Sign-in/Register to view highlights & summary Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
