Anjan Kumar Ray

Robotic Ubiquitous Cognitive Ecology for Smart Homes

Robotic ecologies are networks of heterogeneous robotic devices pervasively embedded in everyday environments, where they cooperate to perform complex tasks. While their potential makes them increasingly popular, one fundamental problem is how to make them both autonomous and adaptive, so as to reduce the amount of preparation, pre-programming and human supervision that they require in real world applications. The project RUBICON develops learning solutions which yield cheaper, adaptive and efficient coordination of robotic ecologies. The approach we pursue builds upon a unique combination of methods from cognitive robotics, machine learning, planning and agent-based control, and wireless sensor networks. This paper illustrates the innovations advanced by RUBICON in each of these fronts before describing how the resulting techniques have been integrated and applied to a proof of concept smart home scenario. The resulting system is able to provide useful services and pro-actively assist the users in their activities. RUBICON learns through an incremental and progressive approach driven by the feedback received from its own activities and from the user, while also self-organizing the manner in which it uses available sensors, actuators and other functional components in the process. This paper summarises some of the lessons learned by adopting such an approach and outlines promising directions for future work.

Decentralized Motion Coordination for a Formation of Rovers

In this paper, a decentralized formation control is proposed which enables collision free coordination and navigation of agents. We present a simple method to define the formation of multi-agents and individual identities (IDs) of agents. Two decentralized coordination and navigation techniques are proposed for the formation of rovers. Agents decide their own behaviors onboard depending upon the motion initiative of the master agent of the formation. In these approaches, any agent can estimate behavior of other agents in the formation. These will reduce the dependency of individual agent on other agents while taking decisions. These approaches reduce the communication burden on the formation where only the master agent broadcasts its motion status per sampled time. Any front agent can act as a master agent without affecting the formation in case of fault in initial master agent. The main idea of this paper is to develop an adequate computational model under which agents in the formation will perform to coordinate among each other. Assignments of IDs to agents are very useful in real-time applications. These proposed schemes are suitable for obstacle avoidance in unknown environment as a whole formation. Agents are free from collision among each other during navigation. These schemes can be used for velocity as well as orientation alignment problems for a multi-agent rover network. These schemes are tested with extensive simulations and responses of agents show satisfactory performances to deal with different environmental conditions.

GPS and sonar based area mapping and navigation by mobile robots

In this paper, we have presented a GPS and sonar based area mapping and navigation scheme for a mobile robot. A mapping is achieved between the GPS space and the world coordinates of the mobile robot which enables us to generate direct motion commands for it. This mapping enables the robot to navigate among different GPS locations within the mapped area. The GPS data is extracted online to get the latitude and longitude information of a particular location. In the training phase, a 2-D axis transformation is used to relate local robot frame with the robot world coordinates and then the actual world coordinates are mapped from the GPS data using a RBFN (radial basis function network) based Neural Network. In the second phase, direct GPS data is used to get the mapping into the world coordinates of mobile robot using the trained network and the motion commands are generated accordingly. The physical placement of sonar devices, their ranging limits and beam opening angles are considered during navigation for possible collision detection and obstacle avoidance. This scheme is successfully implemented in real time with Pioneer mobile robot from ActivMedia Robotics and GPS receiver. The scheme is also tested in the simulation to justify its application in the real world.

Sonar-Based Rover Navigation for Single or Multiple Platforms: Forward Safe Path and Target Switching Approach

In this paper, we have proposed a sensor fusion scheme along with the geometrical modeling of mobile robot navigation path in an unknown environment. In this scheme, the physical placement of sonars, their ranging limits and beam opening angles are considered. A simple 2-D axis transformation is proposed to relate local robot frame with the actual navigation environment. forward safe path (FSP) and target switching approach (TSA) are proposed for efficient obstacle avoidance and target tracking of mobile robot. FSP greatly simplifies the environment conditions as sensed by the robot and also provides minimum turning path during avoidance of obstacles. This method also removes the ldquooscillationrdquo in the mobile robot navigation path. TSA technique gives highest priority on the target tracking during the obstacle avoidance and seeks minimum distance path towards the target. These methods remove unnecessary turning of mobile robot during navigation. A scheme for target directional motion is also proposed. So, mobile robot takes the minimum turning path required towards the target. These methods also ensure the avoidance of ldquodead cycle problemrdquo. These schemes are successfully implemented on a model of PatrolBot mobile robot from ActivMedia Robotics. The overview of current research work on multi-domain robotic system namely system-of-systems is also presented. This paper also describes the Global Positioning System-based navigation of rovers. Results of real-time experiments with Pioneer II P2AT-8 from ActivMedia are included in this paper to show the future aspect of this research work.

Development of Cognitive Capabilities for Smart Home Using a Self-Organizing Fuzzy Neural Network

A smart home requires cognitive assistance to analyze and understand the behavior in this sensory rich environment. In this paper we explore the potential of a self-organizing fuzzy neural network (SOFNN) as a core component of a cognitive system for a smart home environment. We develop a cognitive reasoning module that has the ability to adapt its neuronal structure through adding and pruning of neurons according to the incoming data. The SOFNN rules explore the relations of the inputs and the desired reasoning outputs. The network is trained with realistic synthesized data to show its adaptation capability and is tested with unseen data to validate its cognitive capabilities. We outline the theoretical development and describe the results achieved. This initial implementation of the cognitive module demonstrates the potential of the architecture and will serve as a very important test-bed for future work.

Sonar based Autonomous Automatic Guided Vehicle (AGV) navigation

In this paper, a sonar based navigation scheme for autonomous automatic guided vehicle (AGV) is proposed considering the physical placement, ranging limits and beam opening angles of sonars. A geometrical modeling of navigation path in an unknown environment is presented. A simple 2D axis transformation is proposed to relate local robot frame with the actual environment. Forward safe path (FSP) and target switching approach (TSA) are proposed for efficient obstacle avoidance and target tracking. FSP simplifies the environment conditions and provides minimum turning path during obstacle avoidance. TSA gives highest priority on the target location during obstacle avoidance and generates minimum distance path towards the target, removing unnecessary turning of the AGV. A scheme for target directional motion is proposed to get minimum turning path towards the target. These methods ensure the avoidance of dasiadead cyclepsila problem and generate dasiaoscillationpsila free navigation. These schemes are successfully implemented on a model of PatrolBottrade from ActivMedia Robotics.

Kinematic control of robot manipulators using visual feedback

This paper presents a hybrid visual motor control scheme for robot manipulators using visual feedback. The proposed scheme uses an extended Kohonens Self Organizing Map (EKSOM) to find out the mapping from the task space to the joint space of the manipulator. Given the camera coordinates, the EKSOM has been trained to compute the joint space using visual feedback and system model. This scheme can be used to track the position of a moving object. The position-tracking of moving object is achieved using a prediction rule based on visual feedback from the camera. In the nonredundant case, this scheme is successfully implemented on a three-link manipulator for a known and unknown trajectory. In the redundant case, configuration control is used for the tracking of object position.

Coordinated traffic scheduling for communicating mobile robots

In this paper, a multi-robot networking paradigm is presented. This provides a general framework for coordination among a group of robots. An experiment is conducted showing the effectiveness of the developed network paradigm where a robot controls a group of robots. A coordinated traffic scheduling method is proposed for mobile robots. The aim is to build onboard knowledge for autonomous robots without ranging sensors (sonar or laser range finder) and/or cameras. In this work, more emphasis is given on the exploration of interactions between a pair of robots. The robots share their positions, orientations and safety information and the decision of a robot depends on interactions of the forward safe paths (FSPs) of these robots. The properties of intersection of two straight lines are used to classify different situations. The proposed method is discussed in details with various combinations of scenarios. Simulation results are presented to show the effectiveness of the proposed method.

Inverse Kinematic Control Using Rotational And Joint Space Clustering With Visual Motor Coordination

Abstract In this paper, the inverse kinematic control of a 6-DOF robot manipulator is achieved using visual motor coordination (VMC). Here the positional data is converted into image plane data of a pair of cameras. The Redundancy resolution is a prime goal for the robot manipulator with higher dimensional joint space than the task-space. In this work, we present five schemes for this redundancy resolution based on hybrid visual motor co-ordination (VMC) for a 6-dof robot manipulator by clustering the rotational space and joint space information with visual feedback from a pair of cameras. The proposed schemes are used with the extended Kohonens Self Organizing Map (EKSOM) to find out the mapping from 3-dimensional positional task space to the 6-dimensional joint space of the manipulator. The neural network with EKSOM is modified to use the cyclic nature of angular displacement of joints. The visual feedback is obtained through a pair of calibrated cameras. So, each positional data is converted to corresponding camera coordinates and then the modified EKSOM has been trained to obtain the input-output mapping by combining the visual feedback and hybrid system model consisting of forward kinematics of the manipulator. These methods produce smooth joint movements for positional tracking. These schemes are successfully implemented on a model of 6-DOF PowerCube™ robot manipulator from Amtec Robotics.