Shashi Shekhar Jha

On a Mobile Agent Framework for an Internet of Things

With a large number of heterogeneous devices and gadgets ready to populate the Internet, finding an effective mechanism for communication amongst them has become imperative. Communication between networked heterogeneous devices can allow them to utilize each others individual capabilities to perform a task co-operatively. We propose the use of mobile agents to not only achieve such communication but also to pro-actively search for resources and provide services to the various devices populating such a network. In this paper we describe a mobile agent framework for an Internet of Things and discuss its working through a real implementation.

STIGMERGY-BASED SYNCHRONIZATION OF A SEQUENCE OF TASKS IN A NETWORK OF ASYNCHRONOUS NODES

Social insects are able to build complex architectural structures in spite of their limited capabilities and simple behaviors. The coordination of various activities within a colony of such insects is performed stigmergically through indirect communication. This article describes a stigmergy-based technique by which a set of mobile agents trigger the execution of a sequence of tasks at every node in a network, in a near synchronous manner. These agents use a concept of an innate task-specific potential, which guides them to initiate the execution of a task at various asynchronously operating nodes of a network, in a synchronous manner. This innate potential maintained within the agent also assists in self-healing. The technique is oblivious to both the number of agents and nodes in the network and, hence, is scalable. The proposed technique was emulated on various real static and dynamic topologies by varying different parameters. The results obtained validate the efficacy of the concept of potential. An implementation using real robots has also been presented to portray the practical viability of the proposed technique.

A logic programming interface for multiple robots

Mobile robots have found their way into a range of applications. Writing control programs for these robots is conventionally performed using traditional programming paradigms. Embedding intelligence into such robots calls for an environment that supports logic programming, an aspect that is grossly lacking in the currently available platforms. This paper presents an LPA Prolog based interface to program Lego NXT® robots (LPA-PRO-NXT) that features both concurrent programming of multiple robots as also their control from a remote site. The use of Prolog provides the much needed logic programming support. The paper portrays two real implementations to emphasize the ease of programming concurrency and intelligence for these robots using the developed interface.

Tartarus: a multi-agent platform for integrating cyber-physical systems and robots

The pervasive Internet has lured a variety of embedded devices including sensors nodes, controller boards, smartphones, etc. to form their respective niches within its fold. With a heterogeneous set of devices acting as its nodes, using it as a mere passive entity would be a gross under-utilization of such a massive networked resource. By tethering robots as mobile nodes onto this network, it is possible to realize both the flow of information as also on-demand physical actuation. Managing the communication and control of such a Cyber-Physical System (CPS) is a complex task and calls for an integrated platform that can cater to its scalability, heterogeneity, on-the-fly programming and distributed and decentralized control. In this paper, we introduce a multiagent platform, nicknamed Tartarus, designed to network and integrate robots with other heterogeneous devices to form a CPS. Its inherent agent based technology provides a range of features which include autonomy, intelligence, distributed and decentralized control, among others. The paper describes the platform together with an implementation of a CPS integrated with robots to validate its use in real world scenarios.

On Emulating Real-World Distributed Intelligence Using Mobile Agent Based Localized Idiotypic Networks

Researchers have used Idiotypic Networks in a myriad of applications ranging from function optimization to pattern recognition, learning and even robotics and control. Most of the reported works that have used the Idiotypic network have been simulations wherein not all entities perform in a true distributed, parallel and asynchronous manner. The concentration of an antibody within the network is always assumed to be single valued, which is easily available as a global parameter in such simulated systems. This paper describes a novel architecture and dynamics to emulate an Idiotypic network wherein antibodies within a real physical network interact at antigen-affected nodes, sense their respective global populations stigmergically and form Localized Idiotypic Networks that eventually control their respective global populations across the network. Typhon, a mobile agent platform, running at the various nodes forming the physical network, was used for the emulation. While the mobile agents acted as antibody carriers and ensured their mobility, the nodes forming the physical network formed the antigenic sites. Results, portrayed herein, show the selective rise in global populations of the set of antibodies that are more effective in neutralizing a range of antigens across the network.

Automatic Facial Expression Recognition Using Extended AR-LBP

The Local Binary Pattern (LBP) based operators are sensitive to localization errors. To mitigate these errors input images are manually aligned, face is localized using eyes co-ordinates in the image before feature extraction and multi-scale or multi operators are used, which restricts the use of LBP based operators for automatic facial expression recognition. This paper proposes an Extended Asymmetric Region Local Binary Pattern (EAR-LBP) operator and automatic face localization heuristics to mitigate the localization errors for automatic facial expression recognition. The proposed operator along with face localization heuristics was evaluated for person-independent facial expression recognition on JAFFE and CK+ databases using a multi-class SVM with Linear and Radial Basis Function (RBF) as kernels. It is observed that face localization and the EAR-LBP method are able to mitigate the localization errors to produce reasonably better performance. Maximum 10-fold cross validation average performance of 58.74% and 60.35% were obtained on JAFFE and in case of CK+ database, maximum performance of 83.09% and 82.21% were obtained using Linear and RBF kernels for SVM multi-class classifier respectively.

On a Multi-agent Distributed Asynchronous Intelligence-Sharing and Learning Framework

The current digital era is flooded with devices having high processing and networking capabilities. Sharing of information, learning and adaptation in such highly distributed systems can greatly enhance their performance and utility. However, achieving the same in the presence of asynchronous entities is a complex affair. Multi-agent system paradigms possess intrinsic similarities with these distributed systems and thus provide a fitting platform to solve the problems within. Traditional approaches to efficient information sharing and learning among autonomous agents in distributed environments incur high communication overheads. Non-conventional tactics based on social insect colonies provide natural solutions for transfer of social information in highly distributed and dense populations. This paper portrays a framework to achieve distributed and asynchronous sharing of intelligence and consequent learning among the entities of a networked distributed system. This framework couples localized communication with the available multi-agent technologies to realize asynchronous intelligence-sharing and learning. The framework takes in a user-defined objective together with a learning algorithm as inputs and facilitates cooperative learning among the agents using the mechanisms embedded within. The proposed framework has been implemented using Typhon agent framework over a LAN. The results obtained from the experiments performed using both static and dynamic LANs, substantiate the applicability of the proposed framework in real distributed mobile computing environments.

On Stigmergically Controlling a Population of Heterogeneous Mobile Agents Using Cloning Resource

Cloning can greatly enhance the performance of networked systems that make use of mobile agents to patrol or service the nodes within. Uncontrolled cloning can however lead to generation of a large number of such agents which may affect the network performance adversely. Several attempts to control a population of homogeneous agents and their clones have been made. This paper describes an on-demand population control mechanism for a heterogeneous set of mobile agents along with an underlying application for their deployment as service providers in a networked robotic system. The mobile agents stigmergically sense and estimate the network conditions from within a node and control their own cloning rates. These agents also use a novel concept called the Cloning Resource which controls their cloning behaviour. The results, obtained from both simulation and emulation presented herein, portray the effectiveness of deploying this mechanism in both static and dynamic networks.

On using emotions in decision-making by situated robots

Several attempts have been made to embed emotions into machines so as to improve their abilities and perform better. In this paper, we try to empower a situated robot with a decision-making capability based on dynamics similar to that used to generate synthetic emotions. Using resources made available to it within its environment, the robot selects and performs one of several high and low profile tasks so as to survive comfortably within its environment. The main goal is to ensure that the robot performs as many high profile tasks as possible during its life time. We define a Comfort level that forms a metric to ascertain its usefulness in the environment using which the robot is motivated to perform the tasks. We present the results obtained from actual experiments conducted using a real robot situated in an Artificial-Life-like environment and conclude that the use of emotions in the decision-making process motivates the robot to perform more of the high profile tasks and thus increases its utilization in the given environment.

Orchestrating the Sequential Execution of Tasks by a Heterogeneous Set of Asynchronous Mobile Agents

Coordinated execution of a sequence of tasks by a group of heterogeneous agents possessing different capabilities is known to be a complex problem. Researchers have looked upon nature to capture the mechanisms embedded within the highly distributed swarms of insects to coordinate the execution of tasks by multiple entities. Stigmergy based interactions forms the key that regulates various processes within such swarms. This paper describes a technique by which a heterogeneous set of asynchronous mobile agents comprising a swarm commence the execution of their assigned tasks by coordinating amongst each-other using stigmergy. These mobile agents make use of stimulations from the environment to activate the execution of a sequence of tasks on a network of nodes in a near synchronous manner. A stigmergic cloning controller facilitates the cloning of on-demand agents to provide for parallel executions. The agents also inherently exhibit self-healing. The paper also describes results obtained from emulations performed over a LAN which proves the practical viability of the proposed approach.