Optimal Algorithm Allocation for Single Robot Cloud Systems
Abstract : For a robot to perform a task, several algorithms must be executed, sometimes simultaneously. The algorithms can be executed either on the robot itself or, if desired, on a cloud infrastructure. The term cloud infrastructure refers to hardware, storage, abstracted resources, and network resources associated with cloud computing. Depending on the decision of where the algorithms are executed, the overall execution time and memory required for the robot, change accordingly. The price of a robot depends on its storage capacity and computational power, among other factors. We answer the question of how to maintain a given performance and deploy a cheaper robot (lower resources) by allocating computational tasks to the cloud infrastructure depending on memory, computational power, and communication constraints. Even for a fixed robot, our model provides a way to achieve optimal overall performance. We provide a general model for optimal algorithm allocation decision under certain constraints. We illustrate the model with simulation results. The main advantage of our model is that it provides optimal task allocation simultaneously for memory and time.
? The strategy employed by crossover is to construct new individuals from existing high-performance individuals by recombining sub components .
? Many offloading optimization techniques exist that are adaptable specifically for optimal allocation in CNR.
? Such implementation has resulted in several automated warehouse applications for material handling including conveyers, sorters, goods to picker solutions and other mechanized equipment that has the potential to improve the productivity of the existing workforce.
? Task offloading in multi-robot cloud-assisted systems is multifaceted (contrary to single robot) and more complex, as there exists an added aspect of robot-robot communication along with the robot-cloud communication.
? Our goal is to provide an answer to the allocation problem for a cloud robotic system with a single robot.
? Our methods for finding an optimized allocation of algorithms yield two classes each for memory and time at a given value, with the intersection yielding the optimal solution to minimize the robot’s Memory-Time problem.
? If the intersection has more than one solution to the allocation problem, the decision to the allocation of algorithms can be made by a user depending on the importance of the memory usage or overall completion time factors.
? This work is also a first step towards solving the allocation problem for cloud robotic systems with multiple robots, which goes beyond the above shortcomings of the currently existing solutions.
• We have actively proposed the idea to make movement decisions based on suitable offloading decisions, which essentially points to path planning to accommodate offloading decisions.
• The proposed framework also allows us to analyse the impact of mobility and communication through simulation, for which task offloading is presented as an optimization problem.
• An integrated cloud networked robotics framework is proposed in order to realize both a smart city and smart manufacturing vision while taking into consideration its various complexities.
• Our proposed scenario presents one such multi-objective problem where offloading, path planning and AP selection is jointly considered for making the decisions.
? A dynamic task allocation answers the question of how to achieve the optimal performance of the system by dynamically allocating tasks according to time, (Ti)i?N, in the order of the sets of arrived tasks.
? In other words, static task allocation is not only as important as dynamic task allocation, but it also shows how to achieve the optimal performance of cloud robotic systems while performing any task.
? We have provided a model for users to decide which type of robot is better suited to perform their required tasks, and also how algorithms can be allocated to have the fastest overall performance in the context of cloud robotics.
? Our method provides a solution for achieving optimal performance of the cloud robotic system with a single robot, and also allows comparison between the performances of multiple robots.
|