Philip Asare

MusicalHeart: a hearty way of listening to music

MusicalHeart is a biofeedback-based, context-aware, automated music recommendation system for smartphones. We introduce a new wearable sensing platform, Septimu, which consists of a pair of sensor-equipped earphones that communicate to the smartphone via the audio jack. The Septimu platform enables the MusicalHeart application to continuously monitor the heart rate and activity level of the user while listening to music. The physiological information and contextual information are then sent to a remote server, which provides dynamic music suggestions to help the user maintain a target heart rate. We provide empirical evidence that the measured heart rate is 75% -- 85% correlated to the ground truth with an average error of 7.5 BPM. The accuracy of the person-specific, 3-class activity level detector is on average 96.8%, where these activity levels are separated based on their differing impacts on heart rate. We demonstrate the practicality of MusicalHeart by deploying it in two real world scenarios and show that MusicalHeart helps the user achieve a desired heart rate intensity with an average error of less than 12.2%, and its quality of recommendation improves over time.

Auditeur: a mobile-cloud service platform for acoustic event detection on smartphones

Auditeur is a general-purpose, energy-efficient, and context-aware acoustic event detection platform for smartphones. It enables app developers to have their app register for and get notified on a wide variety of acoustic events. Auditeur is backed by a cloud service to store user contributed sound clips and to generate an energy-efficient and context-aware classification plan for the phone. When an acoustic event type has been registered, the smartphone instantiates the necessary acoustic processing modules and wires them together to execute the plan. The phone then captures, processes, and classifies acoustic events locally and efficiently. Our analysis on user-contributed empirical data shows that Auditeurs energy-aware acoustic feature selection algorithm is capable of increasing the device lifetime by 33.4%, sacrificing less than 2% of the maximum achievable accuracy. We implement seven apps with Auditeur, and deploy them in real-world scenarios to demonstrate that Auditeur is versatile, 11.04% - 441.42% less power hungry, and 10.71% - 13.86% more accurate in detecting acoustic events, compared to state-of-the-art techniques. We present a user study to demonstrate that novice programmers can implement the core logic of interesting apps with Auditeur in less than 30 minutes, using only 15 - 20 lines of Java code.

FSTPA-I: a formal approach to hazard identification via system theoretic process analysis

Cyber-physical systems (CPS) are usually safety critical, making systems safety a CPS issue. Many efforts have been made in safety verification of CPS and some effort has been made in safety-guided design of specific CPS, but fewer efforts have been made in a formal science to aid in safety-guided design. One domain crucial to safety-guided design is hazard analysis, which can be challenging for complex dynamic systems like CPS. Recently, systems theoretic process analysis (STPA) has emerged as a promising hazard analysis technique applicable to CPS; however despite its improvement over traditional techniques, it lacks a solid formal (rig-orous) approach making much of its application ad-hoc and open to a lot of the issues with non-rigorous methods. This paper presents a formal framework for the hazard identification step in STPA (STPA Step One). We show that the formal framework handles many of the issues that arise in a non-rigorous approach and makes the results from analysis less ambiguous and more complete. We also find that an explicit notion of system components is not necessary for undertaking hazard analysis on the system level much in line with the way systems are analyzed in other systems theory fields.

A methodology for developing quality of information metrics for body sensor design

Body sensors networks (BSNs) are emerging technologies that are enabling long-term, continuous, remote monitoring of physiologic and biokinematic information for various medical applications. Because of the varying computational, storage, and communication capabilities of different components in the BSN, system designers must make design choices that trade off information quality with resource consumption and system battery lifetime. Given these trade-offs, there is the possibility that the information presented to the health practitioner at the end point may deviate from what was originally sensed. In some cases, these deviations may cause a practitioner to make a different decision from what would have been made given the original data. Engineers working on such systems typically resort to traditional measures of data quality like RMSE; however, these metrics have been shown in many cases to not correlate well with the notions of information quality for the particular application. Objective metrics of information distortion and its effects on decision making are therefore necessary to help BSN designers make more informed trade-offs between design constraints and information quality and to help practitioners understand the kind of information being produced by BSNs, on which they have to base decisions. In this paper, we present a general methodology for developing such metrics for various BSN applications, illustrate how this methodology can be applied to a real application through a case study, and discuss issues with developing such metrics.

Piecewise Linear Dynamical Model for Action Clustering from Real-World Deployments of Inertial Body Sensors

Human motion has been reported as having great relevance to various disease, disorder, injuries and emotional state. Therefore, motion assessment using inertial body sensor networks (BSNs) is gaining popularity as an outcome measure in clinical study and neuroscience research. The efficacy of motion assessment heavily relies on the accurate temporal clustering of human motion into actions on various time scales. However, two human factors in real-world deployments of inertial BSNs make such motion assessment challenging: mounting errors (where sensor displacement and orientation do not match what is assumed by processing algorithms) and insecure mounting (where sensors are loosely worn causing them to shake during operations). In order to enhance the robustness of human actions clustering from real-world BSN data, this work leverages dynamical systems modeling with the considerations of human factors. By proposing a computational body-model framework called the piecewise linear dynamical model (PLDM), we derive a robust method to segment time series data of inertial BSNs in real-world deployment with human factors into motion primitives and actions. We test the proposed method on three different inertial BSN datasets, extract actions on different temporal scales and recognize the actions into clusters. The experimental results demonstrate the effectiveness of our approach.

RESONATE: reverberation environment simulation for improved classification of speech models

Home monitoring systems currently gather information about peoples activities of daily living and information regarding emergencies, however they currently lack the ability to track speech. Practical speech analysis solutions are needed to help monitor ongoing conditions such as depression, as the amount of social interaction and vocal affect is important for assessing mood and well-being. Although there are existing solutions that classify the identity and the mood of a speaker, when the acoustic signals are captured in reverberant environments they perform poorly. In this paper, we present a practical reverberation compensation method called RESONATE, which uses simulated room impulse responses to adapt a training corpus for use in multiple real reverberant rooms. We demonstrate that the system creates robust classifiers that perform within 5 - 10% of baseline accuracy of non-reverberant environments. We demonstrate and evaluate the performance of this matched condition strategy using a public dataset, and also in controlled experiments with six rooms, and two long-term and uncontrolled real deployments. We offer a practical implementation that performs collection, feature extraction, and classification on-node, and training and simulation of training sets on a base station or cloud service.

Preliminary quantitative evaluation of residential virtual energy storage using power sensing

This work aims to quantify virtual energy storage (VES) monetary cost-savings potential for residential homes. It is part of an effort to develop smart systems (using power sensors, and simple computation and control mechanisms) to assist individuals in making decisions about energy use that will save energy and, consequently, electricity costs. To make a home its own storage system, we need to shrewdly employ the heating, ventilation, and air conditioning (HVAC) system to harness the houses thermal storage capabilities by methods such as preheating or precooling the house during periods when energy is less expensive so that this heat or coolness will be retained during higher-cost periods. This paper presents results from preliminary experimental exploration of the costs and benefits of this approach on a real residential microgrid testbed, through data collection with power sensors and control of the HVAC for the VES case. We also develop a cost-savings model that applies to load management in general to compare (based on the data) VES and battery energy storage (BES) — currently the more traditional and widely-advocated-for approach to energy storage for load management Results indicate better cost-effectiveness of VES (by an order of magnitude in some cases) compared to BES.

Preliminary Exploration of Virtual-Storage-Based Load Management: Poster Abstract

Utilities practice a demand-response, central-generation dispatch model where the amount of energy generated at any instant is controlled to meet customer load. Increases in energy demand and intermittent sources like renewables makes this approach unsustainable. Load management, where demand is adjusted to meet forecasted supply, offers a promising alternative. Storage is essential to this approach. We present preliminary results in our explorations in load management on a residential microgrid testbed using the home as a storage device. A preliminary precool test is carried out on the house. This virtual storage precool strategy is shown to provide considerable cost and energy savings compared to a baseline scenario.

A system for semi-automated management of blood loss during surgery: Preliminary results

This paper presents preliminary results from a pilot study for automated management of blood loss during surgery. The paper describes our approach in developing mathematical models for the cardiovascular system including blood loss and use of colloids for resuscitation. Further, adaptive models are synthesized and used for designing appropriate control algorithms. These algorithms are further integrated into a computerized platform that interfaces with medical devices such as monitors and pumps, which will be tested and validated in animal trials. The eventual development will be minimalist with an open architecture in order to enable easy adoption by others in the community.

BodySim: a multi-domain modeling and simulation framework for body sensor networks research and design

Modeling and simulation play important roles in engineering research and design. These techniques are especially helpful in the early phases where limited detail is available about the design and where design changes are less costly. In addition, high-fidelity models can be employed at the later stages to complement testing. Models are also important research tools for understanding complex phenomena.