As an emerging technology in the era of Industry 4.0, digital twin is gaining unprecedented attention because of its promise to further optimize process design, quality control, health monitoring, decision and policy making, and more, by comprehensively modeling the physical world as a group of interconnected digital models. In a two-part series of papers, we examine the fundamental role of different modeling techniques, twinning enabling technologies, and uncertainty quantification and optimization methods commonly used in digital twins. This second paper presents a literature review of key enabling technologies of digital twins, with an emphasis on uncertainty quantification, optimization methods, open source datasets and tools, major findings, challenges, and future directions. Discussions focus on current methods of uncertainty quantification and optimization and how they are applied in different dimensions of a digital twin. Additionally, this paper presents a case study where a battery digital twin is constructed and tested to illustrate some of the modeling and twinning methods reviewed in this two-part review. Code and preprocessed data for generating all the results and figures presented in the case study are available on GitHub.

Nano-size drones hold enormous potential to explore unknown and complex environments. Their small size makes them agile and safe for operation close to humans and allows them to navigate through narrow spaces. However, their tiny size and payload restrict the possibilities for on-board computation and sensing, making fully autonomous flight extremely challenging. The first step towards full autonomy is reliable obstacle avoidance, which has proven to be technically challenging by itself in a generic indoor environment. Current approaches utilize vision-based or 1-dimensional sensors to support nano-drone perception algorithms. This work presents a lightweight obstacle avoidance system based on a novel millimeter form factor 64 pixels multi-zone Time-of-Flight (ToF) sensor and a generalized model-free control policy. Reported in-field tests are based on the Crazyflie 2.1, extended by a custom multi-zone ToF deck, featuring a total flight mass of 35g. The algorithm only uses 0.3% of the on-board processing power (210uS execution time) with a frame rate of 15fps, providing an excellent foundation for many future applications. Less than 10% of the total drone power is needed to operate the proposed perception system, including both lifting and operating the sensor. The presented autonomous nano-size drone reaches 100% reliability at 0.5m/s in a generic and previously unexplored indoor environment. The proposed system is released open-source with an extensive dataset including ToF and gray-scale camera data, coupled with UAV position ground truth from motion capture.

Product Description Product Description JetHexa is an open source hexapod robot based on Robot Operating System (ROS). It is armed with high-performance hardware, such as NVIDIA Jetson Nano, intelligent serial bus servos, Lidar and HD camera/ 3D depth camera, which can implement robot motion control, mapping and navigation, tracking and obstacle avoidance, custom prowling, human feature recognition, somatosensory interaction and other functions. Adopted novel inverse kinematics algorithm, supporting tripod and ripple gaits and with highly configurable body posture, height and speed, JetHexa will bring user ultimate using experience.JetHexa not only serves as an advanced platform for user to learn and verify hexapod movement, but also provides solutions for ROS development. To help user embark on a new journey of ROS hexapod robotic world, ample ROS and robot learning materials and tutorials are provided. JetHexa is an open source hexapod robot based on Robot Operating System (ROS). It is armed with NVIDIA Jetson Nano, intelligent serial bus servos, Lidar and Monocular camera/ 3D depth camera to implement robot motion control, mapping and navigation, tracking and obstacle avoidance, custom prowling, human feature recognition, somatosensory interaction, etc. Featuring novel inverse kinematics algorithm, tripod and ripple gaits, highly configurable body posture, height and speed, JetHexa will bring user ultimate using experience.JetHexa not only serves as an advanced platform for user to learn and verify hexapod movement, but also provides solutions for ROS development. To help user embark on a new journey of ROS hexapod robotic world, ample ROS and robot learning materials and tutorials are provided. Jetson Nano Control System Jetson Nano Control System NVIDIA Jetson Nano is able to run mainstream deep learning frameworks, such as TensorFlow, PyTorch, Caffe/ Caffe2, Keras, MXNet, and provides powerful computing power for massive AI projects. Powered by Jetson Nano, JetHexa can achieve image recognition, object detection and positioning, pose estimation, semantics segmentation, intelligent analysis and other almighty functions. NVIDIA Jetson Nano is able to run mainstream deep learning frameworks, such as TensorFlow, PyTorch, Caffe/ Caffe2, Keras, MXNet, and provides powerful computing power for massive AI projects. Powered by Jetson Nano, JetHexa can achieve image recognition, object detection and positioning, pose estimation, semantics segmentation, intelligent analysis and other almighty functions. Monocular Camera (with 2DOF Pan-tilt) Monocular camera can rotate up, down, right and left, as well as realize color tracking, autonomous driving and so on. 3D Depth Camera Depth camera can process depth map data, and realize 3D vision mapping navigation. ROS Highlights ROS Highlights 2D Lidar Mapping, Navigation and Obstacle Avoidance JetHexa is loaded with high-performance EAI G4 Lidar that supports mapping with diverse algorithms including Cartographer, Hector, Karto and Gmapping, path planning, fixed-point navigation as well as obstacle avoidance in navigation. RTAB-VSLAM 3D Vision Mapping and Navigation Supporting 3D color mapping in two ways, pure RTAB vision and fusion of vision and Lidar, JetHexa is able to navigate and avoid obstacle in 3D map and execute global relocation. Multi-point Navigation and Obstacle Avoidance Lidar can detect the surroundings in real time, and let JetHexa avoid the obstacles during muti-point navigation. Depth Image Data, Point Cloud Image Through the corresponding API, JetHexa can obtain depth image, color image and point cloud image of the camera. 2D Lidar Mapping, Navigation and Obstacle Avoidance JetHexa is loaded with high-performance EAI G4 Lidar that supports mapping with diverse algorithms including Cartographer, Hector, Karto and Gmapping, path planning, fixed-point navigation as well as obstacle avoidance in navigation. RTAB-VSLAM 3D Vision Mapping and Navigation Supporting 3D color mapping in two ways, pure RTAB vision and fusion of vision and Lidar, JetHexa is able to navigate and avoid obstacle in 3D map and execute global relocation. Multi-point Navigation and Obstacle Avoidance Lidar can detect the surroundings in real time, and let JetHexa avoid the obstacles during muti-point navigation. Depth Image Data, Point Cloud Image Through the corresponding API, JetHexa can obtain depth image, color image and point cloud image of the camera. KCF Target Tracking Based on KCF filtering algorithm, the robot can track the selected target. Depth Camera Obstacle Recognition With the help of depth camera, it can detect the obstacle ahead and pass through the obstacle. Custom Path Prowling User can customize the path and order the robot to prowl along the designed path. Lidar Tracking By scanning the front moving object, Lidar makes robot capable of target tracking. Lidar Guarding Lidar accounts for the role in guarding the surroundings and ringing the alarm when detecting intruder. Color Recognition and Tracking Skilled in color recognition and tracking, the robot can be set to execute different actions according to the colors. Group Control A group of JetHexa can be controlled by only one wireless handle to perform actions uniformly and simultaneously. Intelligent Formation A batch of robots can be controlled to patrol in different formations. Canyon Crossing When Lidar scans the canyon ahead, the robot will adjust its posture and direction to pass through it. Auto Line Following The robot has the ability to recognize the line in color designated by user and prowl following the line. Tag Recognition and Tracking JetHexat is an expert in recognizing and positioning a few AR Tags at the same time. Posture Detection Built-in IMU sensor can detect the body posture in real time. KCF Target Tracking Based on KCF filtering algorithm, the robot can track the selected target. Depth Camera Obstacle Recognition With depth camera, it can detect and pass through the obstacle. Custom Path Prowling User can customize the path and order the robot to prowl along the designed path. Lidar Tracking By scanning the front moving object, Lidar makes robot capable of target tracking. Lidar Guarding Lidar accounts for the role in guarding the surroundings and ringing the alarm when detecting intruder. Color Recognition and Tracking Skilled in color recognition and tracking, the robot can be set to execute different actions according to the colors. Group Control A group of JetHexa can be controlled by only one wireless handle to perform actions uniformly and simultaneously. Tag Recognition and Tracking JetHexat is an expert in recognizing and positioning a few AR Tags at the same time. Canyon Crossing When Lidar scans the canyon ahead, the robot will adjust its posture and direction to pass through it. Auto Line Following The robot has the ability to recognize the line in color designated by user and prowl following the line. Intelligent Formation A batch of robots can be controlled to patrol in different formations. Posture Detection Built-in IMU sensor can detect the body posture in real time. Upgraded Inverse Kinematics Algorithm Upgraded Inverse Kinematics Algorithm One-click Gait Switching One-click Gait Switching JetHexa supports switching between tripod gait and ripple gait at will. JetHexa supports switching between tripod gait and ripple gait at will. "Moonwalk" in Fixed Speed and Height Through inverse kinematics algorithm, JetHexa can maintain stable during SLAM mapping, and moonwalk in a constant speed. Pitch Angle and Roll Angle Adjustment Highly configurable body posture, center of gravity, pitch angle and roll angle enables the hexapod robot to overcome all type of complicated terrains. Direction, Speed, Height and Stride Adjustment JetHexa can make turn and change lane as moving, and support stepless adjustment in linear velocity, angular velocity, stance, height and stride. Body Self-balancing Body Self-balancing The built-in IMU sensor is in charge of detecting the body posture in real time so as to arrange for the robot to adjust its joints to balance the body. Deep Learning and Model Training for AI Creativity Adopting GoogLeNet, Yolo, mtcnn and other neural networks, JetHexa masters deep learning to train models. Through loading various models, it can recognize the targets quickly so as to implement complex AI projects, including waste sorting, mask identification, emotion recognition, etc. Waste Sorting Quick to recognize different waste cards, and place them in the corresponding area in terms of the category. Mask Identification With strong computing power, JetHexa’s AI function can be expanded through deep learning. Emotion Recognition JetHexa is able to recognize facial features accurately to catch every nuance of expression. Deep Learning and Model Training for AI Creativity Adopting GoogLeNet, Yolo, mtcnn and other neural networks, JetHexa masters deep learning to train models. Through loading various models, it can recognize the targets quickly so as to implement complex AI projects, including waste sorting, mask identification, emotion recognition, etc. Waste Sorting Quick to recognize different waste cards, and place them in the corresponding area in terms of the category. Mask Identification With strong computing power, JetHexa’s AI function can be expanded through deep learning. Emotion Recognition JetHexa is able to recognize facial features accurately to catch every nuance of expression. MediaPipe Development, Upgraded AI Interaction MediaPipe Development, Upgraded AI Interaction Based on MediaPipe framework, JetHexa can carry out human body tracking, hand detection, posture detection, overall detection, face detection, 3D detection and more. Based on MediaPipe framework, JetHexa can carry out human body tracking, hand detection, posture detection, overall detection, face detection, 3D detection and more. Fingertip Trajectory Control Fingertip Trajectory Control Human Posture Control Human Posture Control Gesture Recognition Gesture Recognition 3D Face Detection 3D Face Detection ROS Robot Operating System Global Popular Robotic Communication Framework Global Popular Robotic Communication Framework ROS is an open-source meta operating system for robots. It provides some basic services, such as hardware abstraction, low-level device control, implementation of commonly used functionality, message-passing between processes, and package management. And it also offers the tools and library functions needed to obtain, compile, write, and run code across computers. It aims at providing code reuse support for robotics research and development. ROS is an open-source meta operating system for robots. It provides some basic services, such as hardware abstraction, low-level device control, implementation of commonly used functionality, message-passing between processes, and package management.And it also offers the tools and library functions needed to obtain, compile, write, and run code across computers. It aims at providing code reuse support for robotics research and development. Gazebo Simulation Gazebo Simulation JetHexa employs ROS framework and supports Gazebo simulation. Gazebo brings a fresh approach for you to control JetHexa and verify the algorithm in simulated environment, which reduces experimental requirements and improves efficiency. JetHexa employs ROS framework and supports Gazebo simulation. Gazebo brings a fresh approach for you to control JetHexa and verify the algorithm in simulated environment, which reduces experimental requirements and improves efficiency. Body Control Simulation Verify the kinematics algorithm in simulation so as to avoid the damage to the robot due to the algorithm error. Visual Data Visual data is provided for the observation of the robot end and trajectory of the center of gravity to optimize the algorithm. Various Control Methods Various Control Methods WonderAi APP Map Nav APP (Android Only) PC Software Wireless Handle Product Structure EAI G4 Lidar Intelligent Serial Bus Servo Feature 35KG torque, high accuracy, data feedback, easy wiring, 12V supply voltage and strong power. OLED Display Display the controller properties and battery voltage in real time, and supports custom setting. Anodized Metal Bracket Robot’s metal bracket is finely anodized for delicate appearance and long service life. Monocular Camera/ Depth Camera Monocular Camera/ Depth Camera Either 8 megapixel wide-angle monocular Sony camera or Orbbec 3D binocular structured light depth camera can achieve multi-scenario high-accuracy AI recognition. And depth camera can realize 3D mapping and navigation. Multi-functional Expansion Board Multi-functional Expansion Board The onboard IMU sensor can detect the body posture in real time. There are 2-channel PWM servo port, 2 keys, 1 LED, 2 GPIO expansion interfaces and 2 IIC interfaces on the expansion board. JetHexa Parameter 3D Depth Camera Parameter 3D Depth Camera Parameter EAI G4 Lidar Parameter EAI G4 Lidar Parameter Monocular Camera Parameter Monocular Camera Parameter JetHexa Standard Kit JetHexa Advanced Kit Specifications Item Specification JetHexa Parameter Product weight: 2.5 kgMaterial: Full-metal hard aluminum alloy bracket (anodized)Monocular camera pan-tilt: 2 DOF Battery: 11.1V 3500mAh 5C Lipo batteryBattery life: 60minRobot DOF: 18DOFHardware: ROS controller and ROS expansion boardOperating system: Ubuntu 18.04 LTS + ROS MelodicSoftware: PC software + iOS/ Android APPCommunication method: USB/ Wi-Fi/ EthernetProgramming language: Python/ C/ C++/ JavaScriptStorage: 32GB TF cardServo: HX-35H itelligent serial bus servoControl method: Computer/ phone/ handle controlPackage size: 387*356*210mm(length*width*height)Weight (with package): 3.6 kg Battery Parameter Model: 11.1V 3500mAh 5C Lipo batteryCapacity: 3500mAhRated discharge current: 5CPlug: SM plug + DC femaleVoltage: 11.1VSize: 72*55*19 mmWeight: 159gCharger: 12.6V

SAP Integrated Business Planning IBP S4/HANA Functional Lead REMOTE Duration : 6+ months MUST work CST time zone • 2+ years of SAP S4 experience • 10+ years of experience configuring deploying and managing SAP ERP with a focus on Demand Planning • IT Expertise in demand planning and demand…

When a large number of robots try to reach a common area, congestions happen, causing severe delays. To minimise congestion in a robotic swarm system, traffic control algorithms must be employed in a decentralised manner. Based on strategies aimed to maximise the throughput of the common target area, we developed two novel algorithms for robots using artificial potential fields for obstacle avoidance and navigation. One algorithm is inspired by creating a queue to get to the target area (Single Queue Former -- SQF), while the other makes the robots touch the boundary of the circular area by using vector fields (Touch and Run Vector Fields -- TRVF). We performed simulation experiments to show that the proposed algorithms are bounded by the throughput of their inspired theoretical strategies and compare the two novel algorithms with state-of-art algorithms for the same problem (PCC, EE and PCC-EE). The SQF algorithm significantly outperforms all other algorithms for a large number of robots or when the circular target region radius is small. TRVF, on the other hand, is better than SQF only for a limited number of robots and outperforms only PCC for numerous robots. However, it allows us to analyse the potential impacts on the throughput when transferring an idea from a theoretical strategy to a concrete algorithm that considers changing linear speeds and distances between robots.

We present a motion planning algorithm for a class of uncertain control-affine nonlinear systems which guarantees runtime safety and goal reachability when using high-dimensional sensor measurements (e.g., RGB-D images) and a learned perception module in the feedback control loop. First, given a dataset of states and observations, we train a perception system that seeks to invert a subset of the state from an observation, and estimate an upper bound on the perception error which is valid with high probability in a trusted domain near the data. Next, we use contraction theory to design a stabilizing state feedback controller and a convergent dynamic state observer which uses the learned perception system to update its state estimate. We derive a bound on the trajectory tracking error when this controller is subjected to errors in the dynamics and incorrect state estimates. Finally, we integrate this bound into a sampling-based motion planner, guiding it to return trajectories that can be safely tracked at runtime using sensor data. We demonstrate our approach in simulation on a 4D car, a 6D planar quadrotor, and a 17D manipulation task with RGB(-D) sensor measurements, demonstrating that our method safely and reliably steers the system to the goal, while baselines that fail to consider the trusted domain or state estimation errors can be unsafe.

The security issue of mobile robots has attracted considerable attention in recent years. In this paper, we propose an intelligent physical attack to trap mobile robots into a preset position by learning the obstacle-avoidance mechanism from external observation. The salient novelty of our work lies in revealing the possibility that physical-based attacks with intelligent and advanced design can present real threats, while without prior knowledge of the system dynamics or access to the internal system. This kind of attack cannot be handled by countermeasures in traditional cyberspace security. To practice, the cornerstone of the proposed attack is to actively explore the complex interaction characteristic of the victim robot with the environment, and learn the obstacle-avoidance knowledge exhibited in the limited observations of its behaviors. Then, we propose shortest-path and hands-off attack algorithms to find efficient attack paths from the tremendous motion space, achieving the driving-to-trap goal with low costs in terms of path length and activity period, respectively. The convergence of the algorithms is proved and the attack performance bounds are further derived. Extensive simulations and real-life experiments illustrate the effectiveness of the proposed attack, beckoning future investigation for the new physical threats and defense on robotic systems.

This paper improves the performance of RRT*-like sampling-based path planners by combining admissible informed sampling and local sampling (i.e., sampling the neighborhood of the current solution). An adaptive strategy that accounts for the cost progression regulates the trade-off between exploration (admissible informed sampling) and exploitation (local sampling). The paper proves that the resulting algorithm is asymptotically optimal. Furthermore, its convergence rate is superior to that of state-of-the-art path planners, such as Informed-RRT*, both in simulations and manufacturing case studies. An open-source ROS-compatible implementation is also released.

If we give a robot the task of moving an object from its current position to another location in an unknown environment, the robot must explore the map, identify all types of obstacles, and then determine the best route to complete the task. We proposed a mathematical model to find an optimal path planning that avoids collisions with all static and moving obstacles and has the minimum completion time and the minimum distance traveled. In this model, the bounding box around obstacles and robots is not considered, so the robot can move very close to the obstacles without colliding with them. We considered two types of obstacles: deterministic, which include all static obstacles such as walls that do not move and all moving obstacles whose movements have a fixed pattern, and non-deterministic, which include all obstacles whose movements can occur in any direction with some probability distribution at any time. We also consider the acceleration and deceleration of the robot to improve collision avoidance.

AI Planning, Machine Learning and Process Mining have so far developed into separate research fields. At the same time, many interesting concepts and insights have been gained at the intersection of these areas in recent years. For example, the behavior of future processes is now comprehensively predicted with the aid of Machine Learning. For the practical application of these findings, however, it is also necessary not only to know the expected course, but also to give recommendations and hints for the achievement of goals, i.e. to carry out comprehensive process planning. At the same time, an adequate integration of the aforementioned research fields is still lacking. In this article, we present a research project in which researchers from the AI and BPM field work jointly together. Therefore, we discuss the overall research problem, the relevant fields of research and our overall research framework to automatically derive process models from executional process data, derive subsequent planning problems and conduct automated planning in order to adaptively plan and execute business processes using real-time forecasts.