This paper introduces a new modular approach to robotic grasping that allows for finding a trade off between a simple gripper and more complex human like manipulators. The modular approach to robotic grasping aims to understand human grasping behavior in order to replicate grasping and skilled in-hand movements with an artificial hand using simple, robust, and flexible modules. In this work, the design of modular grasping devices capable of adapting to different requirements and situations is investigated. A novel algorithm that determines effective modular configurations to get efficient grasps of given objects is presented. The resulting modular configurations are able to perform effective grasps that a human would consider "stable". Related simulations were carried out to validate the efficiency of the algorithm. Preliminary results show the versatility of the modular approach in designing grippers.

Abstract-This paper presents modular robots for grasping manipulation with locomotion capability. At the beginning, a brief overview on hyper-redundant and modular grasping approach is given. The innovations of this research lie in two points. Firstly, different grasping modes are integrated based on modular approach. Then manipulation capability of robotic arms and flexible locomotion mobility of mobile robots are combined in our current project. Furthermore, the approach in the paper is not only considering manipulation and mobility in modular robots but also, the exploitation of a task priority based approach introduced to manage the trade-off between these two functionalities. Related kinematics of our approach is present systematically. A rational simulation is also given to confirm the idea. In the end a conclusion is given and future work is outlined.

In automatic driving, the recognition of Front-Vehicle taillights plays a key role in predicting the intentions of the vehicle ahead. In order to accurately identify the Front-Vehicle taillights, we first analyze the different characteristics of the vehicle taillight signal, and then propose an improved taillight recognition model based on YOLOv5s. First, CA(coordinate attention) is inserted into the backbone network of YOLOv5s model to improve small target recognition and reduce interference from other light sources; Then, the EIOU Loss is used to solve the class imbalance problem; Finally, EIOU-NMS is used to solve the problem of anchor box error suppression in the recognition process. We use the actual scene video and vehicle taillights dataset to conduct ablation experiments to verify the effectiveness of the improved algorithm. The experimental results show that the mAP value of the model is 9.2% higher than YOLOv5s.INDEX TERMS Autonomous driving, vehicle taillights recognition, ablation experiment.