Abstract

Industrial robots are most often position controlled and insensitive to external forces. In many robotic applications, however, such as teleoperation, haptics for virtual reality, and collaborative robotics, a close cooperation between humans and robots is required. For such applications, force sensing and control capabilities are required for stable interactions with the operator and environment. The robots must also be backdrivable, i.e., the robot must be able to follow user’s induced movements with the least possible resistance. High force efficiency is also desirable. These requirements are different from the design drivers of traditional industrial robots and call for specific actuators and reducers. Many such devices were proposed in the literature. However, they suffer from several drawbacks, offering either a limited reduction ratio or being complex and bulky. This paper introduces a novel solution to this problem. A new differential cable drive reducer is presented. It is backdrivable, has a high efficiency, and a potentially infinite reduction ratio. A prototype actuator using such a reducer has been developed and integrated on a test bench. The experimental characterization of its performance confirms its theoretical advantages.

Highlights

  • Nowadays robots are still mostly used in industrial environments to perform tedious and repetitive tasks autonomously such as holding, moving and assembling parts [1]

  • Even if a larger cable with a higher stiffness could have been used here, as the reduction ratio no more depends on the input pulley’s diameter as with conventional capstan cable drives, this would increase the bulkiness of the reducer as will we demonstrate hereafter

  • Is only 5.4% of the force capacity of the prototype, which is quite low compared to alternative of the force capacity of the prototype, which is quite low compared to alternative solutions that can be solutions that can be used to design a 61:1 reducer. These results more compact and lighter than cable drive systems). These results demonstrate that the actuator demonstrate that the actuator stiffness, while being lower than anticipated due to a lower stiffness, while being lower than anticipated due to a lower transmission stiffness, remain quite high transmission stiffness, remain quite high compared to usual haptic interfaces and collaborative compared to usual haptic interfaces and collaborative robots

Read more

Summary

Introduction

Nowadays robots are still mostly used in industrial environments to perform tedious and repetitive tasks autonomously such as holding, moving and assembling parts [1]. Welding of a car body in the automotive industry or handling and packaging electronic components or food are good examples. Such industrial robots are position controlled, with focus on precision and repeatability regardless of external perturbations, even if some level of force control is required for assembly tasks. Numerous activities involve complex and non-repetitive tasks which are still performed manually. Such tasks cannot be robotized as it would require expensive solutions which, for small series production, are not competitive compared to human workers, especially in low wage countries. For some tasks requiring reactivity and adaptability, no solution even exist as, despite continuous advancements in sensors and artificial intelligence, humans have unmatched sensing, analysis and decision capabilities

Objectives
Methods
Findings
Conclusion
Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call