When a crane lifts an object, the momentum generated by the movement and external forces like wind can cause the load to swing around. The usual method of controlling the load during pick-up and set-down is with taglines – long ropes attached to the load handled by ground personnel,‘riggers’, to manually pull the load into position. This requires workers to be in proximity of the loads, putting them at potential risk if control of the load is lost.
Tensa Equipment Pty Ltd has been developing the Roborigger, a device that connects between the crane hook and the load, to position and orient the load by remote control using inertial forces applied to a flywheel. By controlling the load remotely, workers would no longer need to be in the vicinity of a moving load. With the hardware, motor and remote control developed, the decision software to control the torque the motor generates to stabilise and position the load needed to be refined to accurately compensate for unwanted load movement.
The sensors in the Roborigger measure the speed, orientation and motor output of the Roborigger itself, and the mass of the hanging load, but not the load’s speed and orientation as no sensors are attached to it. Designing acontrol algorithm for the motor to control the movement of something not measured required the mathematical expertise of the Optimisation and Control Research Group. A mathematical model of the Roborigger was built, incorporating all of the sensor information it collects to define its position and movement in space. A sophisticated control algorithm could then be developed to calculate the optimal torque signal for the motor to self-stabilise its position in real time and move smoothly from an initial to a defined final orientation, compensating all the time for additional external forces created by a swinging load or the wind. The algorithm developed, matched to the existing hardware and software limitations of the Roborigger, effectively became a ‘plug-and-play’ component of the final working prototypes for use on construction sites.
The control algorithm designed for the Roborigger is similar to those that keep drones oriented and stable in the air, and can maintain crane load stability and orientation while shifting and rotating the load in the presence of unpredictable disturbances like wind. It simplifies repetitive rotational and positioning operations, and drastically improves worksite safety by allowing riggers to control and orient loads from a safe distance. Because Roborigger is able to stop a load spinning in the wind, it can also extend the normal window of crane operation into stronger winds than previously possible, improving productivity. Users of Roborigger have praised it for the efficiency improvements that have allowed work to be done more quickly, without taglines and with reduced staff levels. These include the erection of the steelwork structure for Sydney’s Central Station and construction of cruise ships in northern Germany.
- Derick Markwell, Managing Director, Roborigger
Control algorithms to autonomously move devices occur everywhere, from cruise control in cars to flight control of drones, and in other mobile devices. Potential further uses for the control algorithm developed for Roborigger extend to any type of robotic application, where the aim is to stabilise and control a pre-determined movement in an optimal manner within a variable environment. The degree of control is a major feature, and the algorithm could be further extended to very high-precision applications, such as installing turbine blades in power-generating wind turbines, where maintaining positional accuracy to within one degree is critical to installation success.
