Madhavan Shanmugavel

Unified meta-modeling framework using bond graph grammars for conceptual modeling

Existing techniques for developing large scale complex engineering systems are predominantly software based and use Unified Modeling Language (UML). This leads to difficulties in model transformation, analysis, validation, verification and automatic code generation. Currently no general frameworks are available to bridge the concept-code gap rampant in design and development of complex, software-intensive mechatronic systems called cyber-physical systems. To fill this gap and provide an alternative approach to Object Management Groups UML/SysML/OCL combination, we propose: Bond Graph based Unified Meta-Modeling Framework (BG-UMF). BG-UMF is a practical and viable alternative and uses a novel hybrid approach based on model unification and integration. The focus is on conceptual design and development of executable models for large systems. The viability of the framework is demonstrated through an application scenario: conceptual design and development of a navigation and control system for a rotor-craft UAV. We analyze limitations of MDA/UML based frameworks for designing Cyber Physical Systems.We propose BG-UMF as a novel, practical and viable alternative based on Bond Graph Grammars.Design Automation (Concept to Code) is achieved by propagating constraints and causality.Output is an executable model, transformable to platform independent code using available tools.Validation is done using an example-concept design and development of a rotorcraft.

Functional design of a parallel pneumatic drive train for regenerative braking

This paper presents the conceptual design and development of a Pneumatic Regenerative Braking System (PRBS) which specifically targets small to medium sized vehicles powered by petroleum fuels. The idea revolves around recovering energy that is normally lost during the braking process and putting it to better use, such as helping the vehicle accelerate. By lessening the burden of an Internal Combustion Engine (ICE), it is hoped that air pollution can be controlled. A Hardware-in-the-Loop test rig was also designed to simulate and test the concept. Analysis of the test results validates the concept and promises a feasible hybrid technology alternative.

Dynamic risk uncertainties by cooperative sites in path planning of UAVs

Abstract This paper aims at formulation of simple decision maker for path planning of UAVs through the sites of risk uncertainties. The decision maker provides the path planner in finding right fly-through way-point to minimize the risk. The risk level at each site is modelled as a dynamic system. The dynamics are caused by uncertainties to the normal risk level at each site. The uncertainties are contributed by: (i) origin of disturbance, and its spread to other sites through (ii) distance, and (iii) level of cooperation between the sites, and (iv) impact of disturbances on other sites. A decision maker based on Bayesian rule is used to find an intermediate fly-through way-point by minimizing the associated with each site. As an initial phase the dynamic risk uncertainties are generated by cooperative sites in path planning of UAVs.

Fault detection in a rotational system with an eccentric load using sliding mode observer

Vehicle health monitoring system (VHMS) is a part of vehicle health management (VHM) and it is commonly used in various industries: particularly in aerospace and automotive. VHMS provides real time feedbacks, information on life cycle reduction and saves operation cost and time. This requires sensors to gather data for processing through many processes such as fault detection and isolation (FDI) before it is sent for prognosis and diagnosis. This paper presents model-based fault detection & reconstruction of state variables using cascaded Sliding Mode Observers (SMOs). An experimental implementation and validation of the SMO in detecting and isolating actuator and sensor faults accurately. The experimental results agrees with simulated estimated states and hence the cascaded SMO faithfully reconstruct the sensor and actuator faults.

A low-cost underwater navigation system for remotely operated underwater vehicle — Design, system-integration, and implementation

This paper presents a low-cost underwater navigation system for Remotely Operated Underwater Vehicles (ROUVs). The navigation uses an on-board IMU signal together with intermittent GPS signal to provide real-time localization. This GPS-aided Inertial Navigation System (INS) estimates latitude and longitude of the ROUV using Kalman filter. A dedicated navigation software with a GUI is built to simulate the orientation of the ROUV and also is able to acquire near real-time data as it does not require post-data analysis. The software is able to access Google Maps and thus plot the live trajectory. This makes the system user-friendly for people with bare minimum knowledge in mathematics of navigation. The navigation system is first tested on ground where reception of GPS signal is good to ensure proper functioning. This is followed by testing with a custom-built ROUV in swimming pools surrounded by high raise-building where the GPS signal is distorted or poor. The navigation system is able to localize the ROUV with an accuracy of 2 to 3 metres. The system is also designed to store the data on-board as well as on surface to facilitate ease-of-search, and to initialize the search operation in the event of loss of communication.

Design of a low cost parallel pneumatic regenerative braking and acceleration assist for light motor vehicles

Land based vehicles, i.e. cars, make up the main mode of transport on Earth, emitting harmful greenhouse gases due to combustion of fossil fuels. By installing regenerative systems in vehicles, the fuel consumption of the internal combustion engine (ICE) is reduced by capturing the energy usually lost during braking and reusing it to help propel the vehicle. The work described in this paper explores the practicality of a pneumatic regenerative braking system (PRBS). A functional prototype was successfully designed and tested to obtain efficiency data from a prototype.

Motion planning of a bipedal walking robot with leg-mounted ultrasonic sensors — An experimental study

This paper presents motion planning of a biped robot using leg-mounted ultrasonic sensors. An inverted pendulum model was used for modeling and simulation of kinematics. The experiments were conducted using a biped prototype. Each leg of the biped consists of four active joints; two for the hip joint, and one each for the knee and ankle joints. Overall, eight joints are controlled to execute biped locomotion. We used PID controller for damping out initial oscillations and stabilize the initial motion while starting from rest, and a PD controller for joint controls to execute motion planning. Pattern generation schemes for single & double support phases and swing phase were used for walking. In addition to stability, and balance, the primary objective is to simulate and conduct experiments to measure the performance of the biped in achieving two basic locomotions: walking and turning. Two path planning problems were attempted: (i) walking in a straight line and (ii) in a zigzag path. The error between the actual and desired path is used as a performance metric. The biped could achieve the desired motions with reasonable accuracy without tip-off or collide on the fences.

Engineering cyber-physical systems — Mechatronics wine in new bottles?

Cyber-Physical system (CPS), one of the nine key technical areas of Industry 4.0, will transform the way we interact with the physical world. After analysing the current definitions, the common challenges and hurdles faced by system designers, we show that CPS is actually a subset of a mechatronic system and suggest a new definition. We also show how system design has been approached differently by experts from different domains and outline practical difficulties in engineering CPS. As existing CPS design approaches attempt synthesis of design models at signal levels only and do not model energy interactions at physical levels, design synthesis and design automation have become difficult problems. Model-driven development holds the key to overcome design synthesis and end-to-end design automation problems.

OSGi-based, embedded, distributed, telematics framework for flexible service provisioning in cyber-physical production systems

Cyber-Physical Production System (CPSS) and Service Provisioning are two inevitable trends and challenges for factories of the future. Existing literature in the Internet of Things (IoT) landscape detail opportunities and difficulties, but performance data to evaluate the overheads and suitability of services provisioning for industrial applications is not readily available. So, we designed and developed a General Telematic Framework for Rapid Application Development using Service-Oriented Architectural (SOA) principles and then provisioned services used in typical CPSS environments on embedded Raspberry Pi hardware modules. Performance profiling demonstrates the suitability for Industry-grade CPSS applications. Other applications such as Unmanned Aerial Vehicle search and rescue, Vehicular Tracking, and distributed management systems are also possible with OSGi-based, embedded telematics system.

Development of a simple, low-cost autopilot system for multi-rotor UAVs

This paper describes the development of a simple, low-cost, embedded auto pilot for an Unmanned Aerial Vehicle (UAV) and its integration to achieve self-flight capability using commercially available off-the shelf modules and components. The autopilot was developed using Arduino Mega 2560 embedded micro-controller platform and integrated with a DJI Spreading Wings S800 Hexa-copter platform, Wookong M Flight Stabilization Controller, and Sensors. A Ground Control Station (GCS) and a wireless communication system to communicate with the UAV was developed using Xbee modules and tested for realtime communication. GCS sends remote control commands, way-points and navigation plan to the on-board autopilot and receives data streams from the UAV. MATLAB was used for creating the Graphical User Interfaces (GUI) for communication. It is capable of sending commands to the micro-controller and receiving data and state information from the UAV in near-real time and was successfully used for tuning the autopilot. Autopilot development and integration was carried out in two stages: (1) Experiments for estimating the physical parameters and manually tuning the Thrust, Roll, Pitch and Yaw Control parameters using a RC-controller. (2) Coupling the Auto pilot (integrating the Arduino Mega 2560 embedded micro-controller) to the Platform and tuning Roll, Pitch, Yaw and thrust control for stable control manoeuvres. We also discuss some integration issues and report the results.