Frank Mokaya

Polaris: getting accurate indoor orientations for mobile devices using ubiquitous visual patterns on ceilings

Ubiquitous computing applications commonly use digital compass sensors to obtain orientation of a device relative to the magnetic north of the earth. However, these compass readings are always prone to significant errors in indoor environments due to presence of metallic objects in close proximity. Such errors can adversely affect the performance and quality of user experience of the applications utilizing digital compass sensors. In this paper, we propose Polaris, a novel approach to provide reliable orientation information for mobile devices in indoor environments. Polaris achieves this by aggregating pictures of the ceiling of an indoor environment and applies computer vision based pattern matching techniques to utilize them as orientation references for correcting digital compass readings. To show the feasibility of the Polaris system, we implemented the Polaris system on mobile devices, and field tested the system in multiple office buildings. Our results show that Polaris achieves 4.5° average orientation accuracy, which is about 3.5 times better than what can be achieved through sole use of raw digital compass readings.

MyoVibe: vibration based wearable muscle activation detection in high mobility exercises

Skeletal muscles are activated to generate the force needed for movement in most high motion sports and exercises. However, incorrect skeletal muscle activation during these sports and exercises, can lead to sub-optimal performance and injury. Existing techniques are susceptible to motion artifacts, particularly when used in high motion sports (e.g. jumping, cycling, etc.). They require limited body movement, or experts to manually interpret results, making them unsuitable in sports scenarios. This paper presents MyoVibe, a wearable system for determining muscle activation in high motion exercise scenarios. MyoVibe senses muscle vibration signals obtained from a wearable network of accelerometers to determine muscle activation. By modeling the characteristics of muscles and high motion noise using extreme value analysis, MyoVibe can reduce noise due to high mobility exercises. Our system can predict muscle activation with greater than 97% accuracy in isometric low motion exercise cases, up to 90% accuracy in high motion exercises.

MARS: a muscle activity recognition system enabling self-configuring musculoskeletal sensor networks

Poor posture and incorrect muscle usage are a leading cause of many injuries in sports and fitness. For this reason, non-invasive, fine-grained sensing and monitoring of human motion and muscles is important for mitigating injury and improving fitness efficacy. Current sensing systems either depend on invasive techniques or unscalable approaches whose accuracy is highly dependent on body sensor placement. As a result these systems are not suitable for use in active sports or fitness training where sensing needs to be scalable, accurate and un-inhibitive to the activity being performed. We present MARS, a system that detects both body motion and individual muscle group activity during physical human activity by only using unobtrusive, non-invasive inertial sensors. MARS not only accurately senses and recreates human motion down to the muscles, but also allows for fast personalized system setup by determining the individual identities of the instrumented muscles, obtained with minimal system training. In a real world human study conducted to evaluate MARS, the system achieves greater than 95% accuracy in identifying muscle groups.

All your jammers belong to us — Localization of wireless sensors under jamming attack

Accurately determining locations of nodes in mobile wireless network is crucial for a myriad of applications. Unfortunately, most localization techniques are vulnerable to jamming attacks where the adversary attempts to disrupt communication between legitimate nodes in the network. In this paper, we propose an approach to localize a wireless node by using jamming attack as the advantage of the network. Our localization technique is divided into two steps. First, we discover the location of the jammer using power adaptation techniques. Then, we use these properties to extrapolate the locations of jammed nodes. We design a localization protocol using this technique, and demonstrate the feasibility of our mechanism by conducting indoor experiments based on IEEE 802.15.4 wireless nodes. Our result shows that for some situations our mechanism can be used to locate mobile nodes under jamming attack.

Burnout: a wearable system for unobtrusive skeletal muscle fatigue estimation

Skeletal muscles are pivotal for sports and exercise. However, overexertion of skeletal muscles causes muscle fatigue which can lead to injury. Consequently, understanding skeletal muscle fatigue is important for injury prevention. Current ways to estimate exhaustion revolve around self-estimation or inference from such sensors as force sensors, electromyography e.t.c. These methods are not always reliable, especially during isotonic exercises. Toward this end, we present Burnout - a wearable system for quantifying skeletal muscle fatigue in an exercise setting. Burnout uses accelerometers to sense skeletal muscle vibrations. From these vibrations, Burnout obtains a region based feature (R- Feature), in the case of this work, the region mean power frequency (R-MPF) gradient to correlate the sensed vibrations to a known ground truth measure of skeletal muscle fatigue, i.e., Dimitrovs spectral fatigue index gradient. We evaluate Burnout on the biceps and quadriceps of 5 healthy participants through four different exercises, collected in a real world environment. Our results show that by using this R-MPF feature on our real world data set, Burnout is able to reduce the error of estimating the ground truth fatigue index gradient by up to 50% on average compared to using the standard MPF feature.

MARS: a muscle activity recognition system using inertial sensors

We present MARS, a muscle activity recognition system that uses inertial sensors to capture the vibrations of active mus-cles. Specifically, we show how accelerometer data capturing these vibrations in the quadriceps, hamstrings and calf muscles of the human leg, can be leveraged to create muscle vibration signatures. We finally show that these vibration signatures can be used to distinguish these muscles from each other with greater than 85% precision and recall.

Collaborative indoor sensing with the sensorfly aerial sensor network

The SensorFly is a novel, low-cost, miniature (29g) controlled-mobile aerial sensor networking platform. Mobility permits a network of SensorFly nodes, unlike fixed networks, to be autonomous in deployment, maintenance and adapting to the environment, as required for emergency response situations such as fire monitoring or survivor search. We demonstrate the ability of the SensorFly system to collaboratively sense the environment (floor temperature) in a demonstration scenario. The SensorFly nodes are tasked to explore the area and transmit sensed data back to a base station. The system partitions tasks among SensorFly nodes based on their capabilities (location, sensors, energy) to achieve concurrent and faster coverage. The real-time sensor data is presented to the user on a display terminal at the base station.

Design Experiences in Minimalistic Flying Sensor Node Platform through SensorFly

Indoor emergency response situations, such as urban fire, are characterized by dangerous constantly changing operating environments with little access to situational information for first responders. In situ information about the conditions, such as the extent and evolution of an indoor fire, can augment rescue efforts and reduce risk to emergency personnel. Static sensor networks that are pre-deployed or manually deployed have been proposed but are less practical due to need for large infrastructure, lack of adaptivity, and limited coverage. Controlled-mobility in sensor networks, that is, the capability of nodes to move as per network needs can provide the desired autonomy to overcome these limitations. In this article, we present SensorFly, a controlled-mobile aerial sensor network platform for indoor emergency response application. The miniature, low-cost sensor platform has capabilities to self deploy, achieve three-dimensional sensing, and adapt to node and network disruptions in harsh environments. We describe hardware design trade-offs, the software architecture, and the implementation that enables limited-capability nodes to collectively achieve application goals. Through the indoor fire monitoring application scenario, we validate that the platform can achieve coverage and sensing accuracy that matches or exceeds static sensor networks and provide higher adaptability and autonomy.

Acquiring musculoskeletal information in active environments, using a wearable system

Poor body form and incorrect muscle usage are a leading cause of injuries in active sports and exercise. For this reason, monitoring of human muscle activity and skeletal motion i.e., musculoskeletal sensing, is important for alleviating injury. Many current monitoring systems are not conducive for this because they are either too invasive, or unable to mitigate the noise induced in the sensed signal by body motion. Our work focuses on creating a framework for acquiring musculoskeletal information in active environments, using a conducive wearable monitoring system. Upon completion, our work will provide an understanding of how to acquire musculoskeletal information, and propose a hardware implementation as well as body motion noise mitigation techniques.

MyoVibe: Enabling Inertial Sensor-Based Muscle Activation Detection In High-Mobility Exercise Environments

Incorrect muscle activation can lead to sub-optimal performance, muscle imbalance, and eventually bodily injury. Consequently, assessing muscle activation is important for both excelling in exercise, athletics, and professional sports in general. Existing techniques for assessing muscle activation, such as electromyography, are invasive, requiring needles inserted directly into the muscle or electrodes that have considerable placement requirements (shaving, gels, etc.). This makes them unsuitable for active environments. In addition, factors such as body motion noise that results from the high-impact movements encountered in active sports environments easily corrupt sensor data. This compounds the unsuitability of these systems in the sports and exercise arena. As a result, such systems have been explored mostly in clinical rather than sports-based scenarios. We present MyoVibe, a system for sensing and determining muscle activation in high-mobility, high-impact exercise scenarios. MyoVibe senses and interprets multiple muscle vibration signals obtained from a wearable network of accelerometers to determine muscle activation. By utilizing a diverse feature set combined with the simple yet effective motion artifact mitigation technique, MyoVibe can reduce inertial sensor noise in these high-mobility exercises. As a result, MyoVibe can detect muscle activation with greater than 97% accuracy.