Salman Taherian

Separating Movement and Gravity Components in an Acceleration Signal and Implications for the Assessment of Human Daily Physical Activity

Introduction Human body acceleration is often used as an indicator of daily physical activity in epidemiological research. Raw acceleration signals contain three basic components: movement, gravity, and noise. Separation of these becomes increasingly difficult during rotational movements. We aimed to evaluate five different methods (metrics) of processing acceleration signals on their ability to remove the gravitational component of acceleration during standardised mechanical movements and the implications for human daily physical activity assessment. Methods An industrial robot rotated accelerometers in the vertical plane. Radius, frequency, and angular range of motion were systematically varied. Three metrics (Euclidian norm minus one [ENMO], Euclidian norm of the high-pass filtered signals [HFEN], and HFEN plus Euclidean norm of low-pass filtered signals minus 1 g [HFEN+]) were derived for each experimental condition and compared against the reference acceleration (forward kinematics) of the robot arm. We then compared metrics derived from human acceleration signals from the wrist and hip in 97 adults (22–65 yr), and wrist in 63 women (20–35 yr) in whom daily activity-related energy expenditure (PAEE) was available. Results In the robot experiment, HFEN+ had lowest error during (vertical plane) rotations at an oscillating frequency higher than the filter cut-off frequency while for lower frequencies ENMO performed better. In the human experiments, metrics HFEN and ENMO on hip were most discrepant (within- and between-individual explained variance of 0.90 and 0.46, respectively). ENMO, HFEN and HFEN+ explained 34%, 30% and 36% of the variance in daily PAEE, respectively, compared to 26% for a metric which did not attempt to remove the gravitational component (metric EN). Conclusion In conclusion, none of the metrics as evaluated systematically outperformed all other metrics across a wide range of standardised kinematic conditions. However, choice of metric explains different degrees of variance in daily human physical activity.

Profiling energy use in households and office spaces

Energy consumption is largely studied in the context of different environments, such as domestic, corporate, industrial, and public sectors. In this paper, we discuss two environments, households and office spaces, where people have an especially strong impact on energy demand and usage. We describe an energy monitoring system which supports continuous and tailored energy feedback, and assess the level of information (energy awareness) that can be gained from time-series energy profiles. Our studies pointed to similarities between households and office spaces and motivated us to profile energy in the same way for both settings. As result, an individualized energy metric is introduced which assists (a) public sharing of energy use, (b) aggregation and combination of energy use across different environments, and (c) comparison among individuals.

A method to compare new and traditional accelerometry data in physical activity monitoring

The accelerometer devices as traditionally used in the epidemiological field for physical activity monitoring (e.g. Actigraph, Actical, and RT3) provide manufacturer-dependent output values called counts that are computed by obscure and proprietary signal processing techniques. This lack of transparency poses a challenge for comparison of historical accelerometer data in counts with data collected using raw accelerometry in S.I. units — m/s2. The purpose of this study was to develop a method that facilitates the compatibility between both methods through conversion of raw accelerometer output data collected with inertial acceleration sensors into Actigraph counts — the most widely used (de facto standard) device brand in epidemiological studies. The basics of the conversion algorithm were captured from the technical specifications of the Actigraph GT1M. Fine-tuning of the algorithm was achieved empirically under controlled conditions using a mechanical shaker device. A pilot evaluation was carried out through physical activity monitoring in free-living scenarios of 19 adult participants (age: 47 ± 11 yrs, BMI: 25.2 ± 4.1 kg-m−2) wearing both devices. The results show that Actigraph counts estimated by the proposed method explain 94.2% of the variation in Actigraph counts (p < 0.001). The concordance correlation coefficient was 0.93 (p < 0.05). The sensitivity for classifying intensity ranged from 93.4% for light physical activity to 70.7% for moderate physical activity.

Event dissemination in mobile wireless sensor networks

Sensor networks are composed of a large number of densely deployed sensors/actuators. Routing protocols are faced with the challenge of delivering data to sinks through multihop routes, in the presence of energy constrained sensor nodes. We present an energy-aware event dissemination protocol for mobile wireless sensor networks. In our proposed model each sink pro-actively constructs a redundant tree in the network to provide reliable delivery of events in the face of dynamic changes and mobility. Scalability is achieved by minimizing the number of participating nodes on the tree, while maintaining high coverage.

Towards real-time profiling of sprints using wearable pressure sensors

On-body sensor systems for sport are challenging since the sensors must be lightweight and small to avoid discomfort, and yet robust and highly accurate to withstand and capture the fast movements associated with sport. In this work, we detail our experience of building such an on-body system for track athletes. The paper describes the design, implementation and deployment of an on-body sensor system for sprint training sessions. We autonomously profile sprints to derive quantitative metrics to improve training sessions. Inexpensive Force Sensitive Resistors (FSRs) are used to capture foot events that are subsequently analysed and presented back to the coach. We show how to identify periods of sprinting from the FSR data and how to compute metrics such as ground contact time. We evaluate our system using force plates and show that millisecond-level accuracy is achievable when estimating contact times.

Profiling sprints using on-body sensors

This paper describes the design, implementation and deployment of a wireless sensor system for athletes. The system is designed to profile sprints based on input from on-body sensors that are wirelessly connected to a nearby infrastructure. We discuss the choice and use of inexpensive Force Sensitive Resistors (FSRs) to measure foot event timings and provide a detailed analysis of the profiling method used to represent high-level information to the coaches and athletes. In this profiling method, we detect sprinting intervals from high-resolution sensor data, and compute the ground contact times for sprinting performances. We validate our results using force plates and show that the system achieves comparable accuracy in measuring the foot contact times (millisecond accuracy) without the limitations of one or few steps.

Sentient bikes for collecting mobility traces in opportunistic networks

We study the problem of building low-cost city-wide location tracking systems with the intention to provide a platform for large-scale human mobility data collection. We take the city-bikes as our first target, motivated by several social networking applications, and choose Cambridge in the UK as our pilot case. We highlight the main application requirements, discuss the practical design issues, and propose a system architecture based on a hybrid sensor network.

SPS: a middleware for multi-user sensor systems

With the increased realisation of the benefits of studying environmental data, sensor networks are rapidly scaling in size, heterogeneity of data, and applications. In this paper, we present a State-based Publish/Subscribe (SPS) framework for sensor systems with many distributed and independent application clients. SPS provides a state-based information deduction model that is suited to many classes of sensor network applications. State Maintenance Components (SMCs) are introduced that are simple in operation, flexible in placement, and decomposable for distributed processing. Publish/Subscribe communication forms the core messaging component of the framework. SPS uses the decoupling feature of Pub/Sub and extends this across the SMCs to support a more flexible and dynamic system structure. Our evaluation, using real sensor data, shows that SPS is expressive in capturing conditions, and scalable in performance.