My Research

I am now a fifth year graduate student in the Department of Mechanical Engineering at the University of Delaware. I work with Prof. S. K. Agrawal in the Mechanical Systems Lab. The main focus of my research is the control and dynamics of cable suspended robots and machines.

Cable Robot Systems

I work with Dr. S. K. Agrawal in the area of cable robot systems. Part of this research was also conducted with Dr. Andy Sury (now with Pathway Technology). There is a variety of advantages of the cable driven systems over conventional serial and parallel robots. To name a few, a large work volume, an ease of heavy material handling in 6DOF, enhanced crane capabilities (no sway and rotation found in an existing single cable crane), and reconfigurability. There are an equally large number of research effort made in universities and research labs around the world. This project has been funded by NSF and NIST. A cable presents attractive advantages but they are also limited - a cable can not provide compression force on an end-effector. These constraints lead to performance deterioration and even instability, if not properly accounted for in the control design. We proposed cotrol techniques for the constrained cable systems. A laboratory sized 6 DOF spatial robot and 3 DOF planar robt have been fabricated and the theories have been verified by simulations and experiments. In the design, a moving platform (MP) is supported by six cables with winches by a base platform(BP). Through three shuttles mounted on the base, two ropes are routed to a triangular end-effector plate. Each vertex of the triangular end-effector plate has two ropes connecting to it. Our work has shown that this geometry has the largest workspace and the highest global index(GCI) of any other design for the same orientation and ratio of lengths of MP and BP.

The control of the cable-driven robot is through a dSpace 1103 board which does the data acquisition and control of the system. A dSpace ControlDesk acts as a front end interface. The control code is written in MATLAB Simulink and is downloaded on to the dSpace board using Real-Time Workshop. The mechanical design consists of an end-effector, suspended by six cables, and driven by pulleys. Direct-drive servomotors from Kollmorgen, fitted with encoders, drive each cable. Each cable has a force sensor in series to measure the tension being transmitted to the end-effector at any time during motion. All required numerical calculations in the algorithms, including inversion of a 6*6 Jacobian matrix and the linear programming algorithm, are coded in C to achieve real time implementations. In addition, a force control loop using six loadcell sensors is used to compensate for potential sources of error in the experiment due to cable friction in the drive train.

Cargo Transfer Systems - Dual Stage Cable Robot

Skin-to-skin transfer of cargo is a challenging task. This operation in mid sea can be unstable unless the sea motions are properly accoundted for. In a typical skin-to-skin transfer operation, the end-effector and the traget are located on different ships and hense subjected to disturbances, that may not be in phase. NIST has proposed a two-stage cable robot, each stage with six DOF, for this purpose. The upper stage is motivated from keeping the cables away from the adjacent containers, and to increase the system redundancy. The lower stage is designed to engage the container. This two-stage design promises to achieve the desired goal of transfer of cargo from one ship to another safely. We presented the dynamic model for this new type of crane robot, incorporating the disturbance from the sea condition.

(Click the picture to look at in phase motion)

(Click the picture to look at out of phase motion)

(Planar Version of Dual Stage Robot)

Helicopter with Cable Robots

I also worked on dynamics and control of a helicopter carrying a payload using a cable-suspended robot. Transport of externally suspended loads is an important mission of a helicopter. Both military and commercial operators exploit this capability of a helicopter to rapidly move heavy loads to locations where the use of ground based equipment for transport is not possible. Helicopters are actively used for ship replenishment and hosit operations from sea-going cargo vessels, specially in adverse weather. Our work is to enable the dual system to execute tasks that cannot be accomplished by the robots individually. The helicopter is used for gross motion, while the 6DOF cable robot is used for fine manipulation. This robot design promises to be effective for transfer of cargo.

(Click the picture to see movie)

Tether Guided Helicopter Landing

I also worked with Dr. Himanshu Pota and Dr. Matt Garrat, a team in ADFA(Australia Defense Force Academy) to develope Autonomous Helicopter Landing System. The landing of autonomous vehicles is typically attempted using vision and global positioning systems. Vision guided landing uses the assumption that the target's shape is known and the target is moving slowly. A more recent approach for landing of manned helicopter is to use a tether, which is reeled-out from the helicopter as it comes near the deck. The end of the tether is then secured to the deck. The tether is kept taut, and it provides a useful reference for the pilot to make maneuvers and react to the relative motion of the helicopter and the ship. The pilots of Canadian navy and others are using such a protocol to land a helicopter on small-sized ships in rough weather. we addressed the problem ofautonomous landing of a helicopter on a ship's deck using a tether. Although the navy adopts this protocol for landing manned helicopters on ships in rough sea, the unmanned autonomous landing problem using a tether is unexplored.

(Click the picture to see movie)