Jindong Tan

Distributed multi-robot coordination in area exploration

This paper proposes a reliable and efficient multi-robot coordination algorithm to accomplish an area exploration task given that the communication range of each robot is limited. This algorithm is based on a distributed bidding model to coordinate the movement of multiple robots. Two measures are developed to accommodate the limited-range communications. First, the distances between robots are considered in the bidding algorithm so that the robots tend to stay close to each other. Second, a map synchronization mechanism, based on a novel sequence number-based map representation and an effective robot map update tracking, is proposed to reduce the exchanged data volume when robot subnetworks merge. Simulation results show the effectiveness of the use of nearness measure, as well as the map synchronization mechanism. By handling the limited communication range we can make the coordination algorithms more realistic in multi-robot applications.

Heartbeat-Driven Medium-Access Control for Body Sensor Networks

In this paper, a novel time division multiple access based MAC protocol designed for body sensor networks (BSNs) is presented. H-medium-access control (MAC) aims to improve BSNs energy efficiency by exploiting heartbeat rhythm information, instead of using periodic synchronization beacons, to perform time synchronization. Heartbeat rhythm is inherent in every human body and observable in various biosignals. Biosensors in a BSN can extract the heartbeat rhythm from their own sensory data by detecting waveform peaks. All rhythms represented by peak sequences are naturally synchronized since they are driven by the same source, i.e., the heartbeat. Following the rhythm, biosensors can achieve time synchronization without having to turn on their radio to receive periodic timing information from a central controller, so that energy cost for time synchronization can be completely eliminated and the lifetime of the network can be prolonged. An active synchronization recovery scheme is also developed, including two resynchronization approaches. The algorithms are simulated using the discrete event simulator OMNet + + with real-world data from the Massachusetts Institute of Technology-Bostons Beth Israel Hospital multiparameter database Multiparameter Intelligent Monitoring for Intensive Care. The results show that H-MAC can prolong the network life dramatically.

Selection and Navigation of Mobile Sensor Nodes Using a Sensor Network

Hybrid sensor networks comprise of mobile and static sensor nodes setup for the purpose of collaboratively performing tasks like sensing a phenomenon or monitoring a region. In this paper, we present a novel approach for navigating a mobile sensor node (MSN) through such a hybrid sensor network. The static sensor nodes in the sensor network guide the MSN to the phenomenon. One or more MSNs are selected based on their proximity to the detected phenomenon. Navigation is accomplished using the concepts of credit based field setup and navigation force from static sensor nodes. Our approach does not require any prior maps of the environment thus, cutting down the cost of the overall system. The simulation results have verified the effectiveness of the proposed approach. In each of the simulation runs, the static sensor nodes were able to successfully guide the MSN towards the phenomenon

DietCam: Automatic dietary assessment with mobile camera phones

Obesity has become a severe health problem in developed countries, and a healthy food intake has been recognized as the key factor for obesity prevention. This paper presents a mobile phone based system, DietCam, to help assess food intakes with few human interventions. DietCam only requires users to take three images or a short video around the meal, then it will do the rest. The experiments of DietCam in real restaurants verify the possibility of food recognition with vision techniques.

Integrated Task Planning and Control for Mobile Manipulators

In this paper we present an approach to decoupled force/position control of the end-effector along the same direction for redundant robots, and an approach to nonholonomic cart pushing with mobile manipulators. The mobile manipulator is considered as a redundant robot, and a unified dynamic model for an integrated mobile platform and on-board manipulator is developed. The dynamic model is decoupled and linearized using the nonlinear feedback technique in a unified frame. Combining the event-based planning and control method with singularity analysis of the robot arm, a task level action controller is designed and an online kinematic redundancy resolution scheme is developed. The system is stable during normal operation as well as at the occurrence of unexpected obstacles. In addition, explicit force/position control along the same task direction for redundant robots is proposed. The kinematic redundancy of mobile manipulators enables independent control of force and position along the same task directions. To verify the decoupled force/postion scheme, an integrated task planning and control approach is further proposed for the mobile manipulator to complete complicated tasks by regulating its output force. A cart pushing task, which requires both force and position control along the same task direction, is discussed. The cart manipulation task fully integrates trajectory and force planning of the cart, and planning and control of the mobile manipulators. The approaches have been tested on a mobile manipulator consisting of a Nomadic XR4000 and a Puma 560 robot arm. The experimental results demonstrate the efficacy of the approach for the mobile manipulation of a nonholonomic cart.

Sambot: A Self-Assembly Modular Robot System

The design and structure of a self-assembly modular robot (Sambot) are presented in this paper. Each module has its own autonomous mobility and can connect with other modules to form robotic structures with different manipulation abilities. Sambot has a versatile, robust, and flexible structure. The computing platform provided for each module is distributed and consists of a number of interlinked microcontrollers. The interaction and connectivity between different modules is achieved through infrared sensors and Zigbee wireless communication in discrete state and control area network bus communication in robotic configuration state. A new mechanical design is put forth to realize the autonomous motion and docking of Sambots. It is a challenge to integrate actuators, sensors, microprocessors, power units, and communication elements into a highly compact and flexible module with the overall size of 80 mm × 80 mm × 102 mm. The work describes represents a mature development in the area of self-assembly distributed robotics.

Real-time Daily Activity Classification with Wireless Sensor Networks using Hidden Markov Model

This paper presents a hidden Markov model (HMM) approach for real-time activity classification using signals from wearable wireless sensor networks. A wearable wireless sensor network can be used to continuously monitor the daily activities of a subject in real time. However, the wireless sensor nodes are constrained by limited battery and computing resources. The proposed HMM framework has been applied to find the most probable activity states series with low data transmission rate, which makes it highly suitable for daily activity classification applications. The performance was evaluated using a small sensor network consisting of three accelerometers. The activity detection rate is 95.82%, using a test set of 5 subjects with 11 activity series.

Heartbeat driven medium access control for body sensor networks

In this paper, a novel time division multiple access based MAC protocol designed for body sensor networks (BSNs) is presented. H-medium-access control (MAC) aims to improve BSNs energy efficiency by exploiting heartbeat rhythm information, instead of using periodic synchronization beacons, to perform time synchronization. Heartbeat rhythm is inherent in every human body and observable in various biosignals. Biosensors in a BSN can extract the heartbeat rhythm from their own sensory data by detecting waveform peaks. All rhythms represented by peak sequences are naturally synchronized since they are driven by the same source, i.e., the heartbeat. Following the rhythm, biosensors can achieve time synchronization without having to turn on their radio to receive periodic timing information from a central controller, so that energy cost for time synchronization can be completely eliminated and the lifetime of the network can be prolonged. An active synchronization recovery scheme is also developed, including two resynchronization approaches. The algorithms are simulated using the discrete event simulator OMNet ++ with real-world data from the Massachusetts Institute of Technology-Bostons Beth Israel Hospital multiparameter database Multiparameter Intelligent Monitoring for Intensive Care. The results show that H-MAC can prolong the network life dramatically.

An Adaptive-Gain Complementary Filter for Real-Time Human Motion Tracking With MARG Sensors in Free-Living Environments

High-resolution, real-time data obtained by human motion tracking systems can be used for gait analysis, which helps better understanding the cause of many diseases for more effective treatments, such as rehabilitation for outpatients or recovery from lost motor functions after a stroke. In order to achieve real-time ambulatory human motion tracking with low-cost MARG (magnetic, angular rate, and gravity) sensors, a computationally efficient and robust algorithm for orientation estimation is critical. This paper presents an analytically derived method for an adaptive-gain complementary filter based on the convergence rate from the Gauss-Newton optimization algorithm (GNA) and the divergence rate from the gyroscope, which is referred as adaptive-gain orientation filter (AGOF) in this paper. The AGOF has the advantages of one iteration calculation to reduce the computing load and accurate estimation of gyroscope measurement error. Moreover, for handling magnetic distortions especially in indoor environments and movements with excessive acceleration, adaptive measurement vectors and a reference vector for earths magnetic field selection schemes are introduced to help the GNA find more accurate direction of gyroscope error. The features of this approach include the accurate estimation of the gyroscope bias to correct the instantaneous gyroscope measurements and robust estimation in conditions of fast motions and magnetic distortions. Experimental results are presented to verify the performance of the proposed method, which shows better accuracy of orientation estimation than several well-known methods.

Using Bluetooth and sensor networks for intelligent transportation systems

Safety of road travel can be increased if vehicles can be made to form groups for communicating data among themselves. The Bluetooth protocol can be used for inter-vehicle communication among vehicles equipped with Bluetooth devices. This work presents a novel approach to increase the safety of road travel using the concepts of wireless sensor networks and the Bluetooth protocol. We discuss how vehicles can form mobile ad-hoc networks and exchange data sensed by the on-board sensors. The fusion of these data could give a better understanding of the surrounding traffic conditions. The feasibility of using Bluetooth for data exchange among vehicles is evaluated. Coverage area and probability of detection plots for isotropic and non-isotropic sensors are analyzed to study their use to avoid potential dangerous situations in traffic. As the simulation results show, Bluetooth and sensor networks can be used collaboratively to increase the safety of road travel.