Avik Ghose

Feasibility Analysis for Estimation of Blood Pressure and Heart Rate using A Smart Eye Wear

The major issue with hypertension is the fact that it does not have any symptoms also measurements are highly variable in nature. However, high blood pressure has severe effects on a persons health. This poses a requirement for continuous blood pressure monitoring and analysis of people suffering from chronic condition. In this paper we propose PPG based continuous blood pressure and heart rate monitoring system in form factor of a Smart Eye wear. We do a feasibility analysis of the idea and show that it is possible to determine blood pressure and heart rate by sensing PPG from the side of head region (temple) where a spectacle frame would fit. The advantage of Smart Eye Wear over other wearable device is that it provides better contact, minimal motion artifact and maintains uniform pressure without causing any trouble to user.

Unobtrusive indoor surveillance of patients at home using multiple Kinect sensors

In this paper we propose a system for unobtrusive automated indoor surveillance of subjects in indoor environment using the Kinect sensor. We demonstrate that the features of identity, location and activity of a person can be detected with considerable accuracy using the system. Further, we show how existing design patterns can be used to create a data parallel and scalable architecture for such surveillance in real-time.

Exploiting IMU Sensors for IOT Enabled Health Monitoring

Inertial Measurement Units (IMUs) embedded in commercial mobile devices are a good choice for continuous monitoring in healthcare domain due to their attractive form factor and low power consumption. We present improved and accurate sensing algorithms using a single IMU to sense basic events like step count, stride length, fall, immobility etc. Our algorithms have been shown to perform better than the state of the art algorithms, and are implemented in such a way that IMU is not bound to any specific position or orientation with respect to the user. We propose a 3-layer based framework for a complete end-to-end system architecture for IoT enabled health monitoring, useful for application in areas like individual fitness monitoring and elderly care.

Statistical analysis of road-vehicle-driver interaction as an enabler to designing behavioural models

Ƞ Telematics form an important technology enabler for intelligent transportation systems. By deploying on-board diagnostic devices, the signatures of vehicle vibration along with its location and time are recorded. Detailed analyses of the collected signatures offer deep insights into the state of the objects under study. Towards that objective, we carried out experiments by deploying telematics device in one of the office bus that ferries employees to office and back. Data is being collected from 3-axis accelerometer, GPS, speed and the time for all the journeys. In this paper, we present initial results of the above exercise by applying statistical methods to derive information through systematic analysis of the data collected over four months. It is demonstrated that the higher order derivative of the measured Z axis acceleration samples display the properties Weibull distribution when the time axis is replaced by the amplitude of such processed acceleration data. Such an observation offers us a method to predict future behaviour where deviations from prediction are classified as context-based aberrations or progressive degradation of the system. In addition we capture the relationship between speed of the vehicle and median of the jerk energy samples using regression analysis. Such results offer an opportunity to develop a robust method to model road-vehicle interaction thereby enabling us to predict such like driving behaviour and condition based maintenance etc.

Design insights for a mobile based sensor application framework: for aiding platform independent algorithm design

Modern day smart phones are powerful connected sensory and computation nodes for crowd-sensing, urban-sensing and personal-sensing applications. We have developed an Internet of Things (IoT) platform that can seamlessly handle data from the wide variety of sensors available on mobile phones. It can store and run aggregated analysis on the data in real-time. However, mobile phones themselves are a very heterogeneous set of devices. Each phone comes with a different array of sensors with varying sensitivity and control functions. Also, there are multiple development environments and programming languages. A final problem is seamless prototyping of applications offline and then seamless partitioning of the algorithm between phone and the cloud. In this paper we present early design elements of a framework aimed at addressing these issues.

Shake meter: An Autonomous Vibration Measurement System using Optical Strobing: Demo Abstract

In this paper, we intend to demonstrate a novel system to measure the high-speed vibration of an anonymous vibrating object using COTS camera and optical strobing. The whole process is unobtrusive and frugal, can be used in machine inspection. The camera used has a frame rate of 30 frames per second (fps), so in conventional fashion, it is incapable to detain significant information about any vibration frequency which is not in the range of Nyquist theory of frequency (within a range of ±15 Hz). We have solved the challenge using optical strobing phenomena for capturing modulo (of division) between objects frequency and strobing frequency using camera. Motion in the video is tracked by conventional image processing technique. Finally, object vibration is calculated from the frequency plot and optical strobing frequency. Under most of the cases, the application estimates vibration frequency values, within a range of ±1.5% of error.

Mobile sensing framework for task partitioning between cloud and edge device for improved performance

Recently smartphones are used every area in day-to-day life. Smartphones comes with several built-in sensors like gyroscope, accelerometer etc., along with powerful processing units. There exist various frameworks which use mobile as sensing device and mobile sensors as data extractor and process extracted data to calculate various parameter. This processing unit can be resided either in mobile side or cloud side, which provides flexibility to the researcher/developer to reduce computation time by migrating processing unit and transferring data to the cloud side. This may create problem of packet dropping or network issue while transferring data to the cloud. To overcome network issue, we propose a common framework which maintains trade-off between network overhead and processing time. The key feature of proposed framework is dividing processing unit into mobile and cloud side, sends raw data to cloud after preprocessing at mobile side. This will take very low processing time and reduce raw data size, which reduces number of packets to send to the cloud. We investigate feasibility of our proposed framework by implementing and testing with several collaborative sensing applications and comparing with the existing framework. Our result shows promising result by trading off between on-board processing and network overhead across all the solutions we had tested.

Demo: Exploiting IMU Sensors for IoT Enabled Health Monitoring

Inertial Measurement Units (IMUs) embedded in commercial mobile devices are a good choice for continuous monitoring in healthcare domain due to their attractive form factor and low power consumption. We present improved and accurate sensing algorithms to sense basic events like step count, stride length, fall, and calorie, with accuracies better than those of existing arts. The events can be directed to a server running event analytics, yielding important cues about subjects health. One of the important spheres is elderly health- care where such a system might promote a better insight to the subjects health parameters. Fall Detection. For an improved fall detection, we propose the algorithm as in Figure 1, which shows a substantial improvement on mobifall dataset [3] compared to existing arts, with an average sensitivity of 0:855 and a more robust false event rejections. The detection can be used to trigger an automatic alarm for the need of an urgent attention to the elderly subject.

Table of interest: activity recognition and behaviour analysis using a battery less wearable sensor

Energy overheads continue to be a major impediment for wearable based activity recognition systems. We proposed a hybrid approach, which combines wearable-based human sensing with object interaction tracking, for robust detection of ADLs in smart homes. Our proposed framework includes: (a) battery less, low sampling rate, wearable RF sensor tags, that are powered intermittently by an RFID reader, and (b) additional passive RF tags, mounted on daily use objects, that capture the presence and use of specific objects while performing such ADLs. Using an initial experimental set up, we show the ability to recognize activities like eating, typing and reading, which are generally performed on a table, with an accuracy of 96%. Moreover, by capturing the item-level interactions of a user while performing ADLs, this approach can help observe the evolution of fine-grained behavioral changes and anomalies in an individual.

Analysing Multi-Point Multi-Frequency Machine Vibrations using Optical Sampling

Vibration analysis is a key troubleshooting methodology for assessing the health of factory machinery. We propose an unobtrusive framework for at-a-distance visual estimation of such (possibly high frequency) vibrations, using a low fps (frames-per-second) camera that may, for example, be mounted on a workers smart-glass. Our key innovation is to use an external stroboscopic light source (that, for example, may be provided by an assistive robot), to illuminate the machine with multiple mutually-prime strobing frequencies, and use the resulting aliased signals to efficiently estimate the different vibration frequencies via an enhanced version of the Chinese Remainder Theorem. Experimental results show that our technique estimates multiple such frequencies faster, and compares favourably to an equipment-mounted accelerometer alternative, with frequency estimation errors below 0.5% for vibrations occurring up to 500 Hz.