Mobile Robots and Cobots Integration: A Preliminary Design of a Mechatronic Interface by Using MBSE Approach

Enabling technologies that drive Industry 4.0 and smart factories are pushing in new equipment and system development also to prevent human workers from repetitive and non-ergonomic tasks inside manufacturing plants. One of these tasks is the order-picking which consists in collecting parts from the warehouse and distributing them among the workstations and vice-versa. That task can be completely performed by a Mobile Manipulator that is composed by an industrial manipulator assembled on a Mobile Robot. Although the Mobile Manipulators implementation brings advantages to industrial applications, they are still not widely used due to the lack of dedicated standards on control and safety. Furthermore, there are few integrated solutions and no specific or reference point allowing the safe integration of mobile robots and cobots (already owned by company). This work faces the integration of a generic mobile robot and collaborative robot selected from an identified set of both systems. The paper presents a safe and flexible mechatronic interface developed by using MBSE principles, multi-domain modeling, and adopting preliminary assumptions on the hardware and software synchronization level of both involved systems. The interface enables the re-using of owned robot systems differently from their native tasks. Furthermore, it provides an additional and redundant safety level by enabling power and force limiting both during cobot positioning and control system faulting.

Actuation-Coordinated Mobile Parallel Robots With Hybrid Mobile and Manipulation Function

Mobile robots with manipulation capability are a key technology that enables flexible robotic interactions, large area covering and remote exploration. This paper presents a novel class of actuation-coordinated mobile parallel robots (ACMPRs) that utilize parallel mechanism configurations and perform hybrid moving and manipulation functions through coordinated wheel actuators. The ACMPRs differ with existing mobile manipulators by their unique combination of the mobile wheel actuators and the parallel mechanism topology through prismatic joint connections. The common motion of the wheels will provide the mobile function while their differentiation will actuate the parallel manipulator function. This new concept reduces the actuation requirement and increases the manipulation accuracy and mobile motion stability through the coordinated and connected wheel actuators comparing with existing mobile parallel manipulators. The relative wheel location on the base frame also enables a reconfigurable base size with variable moving stability on the ground. The basic concept and general type synthesis are introduced and followed by the kinematics and inverse dynamics analysis of a selected three limb ACMPR. A numerical simulation also illustrates the dynamics model and the motion property of the new mobile parallel robot. The work provides a basis for introducing this new class of robots for potential applications in surveillance, industrial automation, construction, transportation, human assistance, medical applications and other operations in extreme environment such as nuclear plants, Mars, etc.

TANGO: Commonsense Generalization in Predicting Tool Interactions for Mobile Manipulators

Robots assisting us in factories or homes must learn to make use of objects as tools to perform tasks, e.g., a tray for carrying objects. We consider the problem of learning commonsense knowledge of when a tool may be useful and how its use may be composed with other tools to accomplish a high-level task instructed by a human. We introduce TANGO, a novel neural model for predicting task-specific tool interactions. TANGO is trained using demonstrations obtained from human teachers instructing a virtual robot in a physics simulator. TANGO encodes the world state consisting of objects and symbolic relationships between them using a graph neural network. The model learns to attend over the scene using knowledge of the goal and the action history, finally decoding the symbolic action to execute. Crucially, we address generalization to unseen environments where some known tools are missing, but alternative unseen tools are present. We show that by augmenting the representation of the environment with pre-trained embeddings derived from a knowledge-base, the model can generalize effectively to novel environments. Experimental results show a 60.5-78.9% improvement over the baseline in predicting successful symbolic plans in unseen settings for a simulated mobile manipulator.