16 research outputs found
Artificial Intelligence and Machine Learning
No full textWith the ever-expanding digitalisation, more information or data is generated and is available within the digital ecosystem. The expansion of available data and the increased competition caused by globalisation contribute to why manufacturers are looking for more advanced methods to optimise their production and products. The general usage of artificial intelligence (AI) within different fields is expanding. As part of Industry 4.0, AI is also gaining interest within the industrial sector, where companies are expanding and trying different usages of AI, both within their production and as a product or service
Artificial intelligence and internet of things in small and medium-sized enterprises:A survey
No full textInternet of things (IoT) and artificial intelligence (AI) are popular topics of Industry 4.0. Many publications regarding these topics have been published, but they are primarily focused on larger enterprises. However, small and medium-sized enterprises (SMEs) are considered the economic backbone of many countries, which is why it is increasingly important that these kinds of companies also have easy access to these technologies and can make them operational. This paper presents a comprehensive survey and investigation of how widespread AI and IoT are among manufacturing SMEs, and discusses the current limitations and opportunities towards enabling predictive analytics. Firstly, an overview of the enablers for AI and IoT is provided along with the four analytics capabilities. Hereafter a comprehensive literature review is conducted and its findings showcased. Finally, emerging topics of research and development, making AI and IoT accessible technologies to SMEs, and the associated future trends and challenges are summarised.</p
Concept of Easy-to-use Versatile Artificial Intelligence in Industrial Small & Medium-sized Enterprises
Full text linkIn this paper, the concept of what we call AI-Box is presented. This concept is targeting small and medium-sized enterprises within the manufacturing industry sector. The AI-Box aims to bring technologies from Industry 4.0 to them, with the use of easy-to-use and versatile implementation. Preliminary experiments have been conducted at Aalborg University and at an industrial partner to solve vision tasks, which would be too expensive with conventional vision techniques. Moreover, three different convolutional neural networks were tested to find the best- suited architecture. The three networks tested were the simple AlexNet, the complex ResNeXt, and small and complex SqueezeNet. Our results show that it is possible to solve the classification problem in a few epochs. Furthermore, with the use of augmented data, the performance can be improved. Our preliminary results also showed that the simpler convolutional neural network architecture from AlexNet yields a better result when classifying simple data
Robot Learning From a Human Expert Using Inverse Reinforcement Learning: A Deep Reinforcement Learning Approach for Industrial Applications
No full textThe need for adaptable models, e.g. reinforcement learning (RL), have in recent years been more present within the industry. However, the number of commercial solutions using RL is limited, one reason being the complexity related to the design of RL. Therefore, a method to identify complexities of RL for industrial applications is presented in this thesis. It was used on 15 applications inspired from four industrial companies. Complexity was especially identified in relation to the reward functions. Thus two Linear Inverse RL (IRL) algorithms in which the reward function is represented as a linear combination of features, was tested using expert data. Some of the tests indicated a visual better result than tests carried out using RL. The process of designing features shared similarities with the process of designing a reward function. The added complexity of implementing Linear IRL and constructing expert data is thus not always a simpler approach. The IRL method GAIL, which requires no feature construction, was furthermore tested showing potential
Transferring Human Manipulation Knowledge to Robots with Inverse Reinforcement Learning
Full text linkThe need for adaptable models, e.g. reinforcement learning, have in recent years been more present within the industry. In this paper, we show how two versions of inverse reinforcement learning can be used to transfer task knowledge from a human expert to a robot in a dynamic environment. Moreover, a second method called Principal Component Analysis weighting is presented and discussed. The method shows potential in the use case but requires some more research
A data-driven modular architecture with denoising autoencoders for health indicator construction in a manufacturing process
No full textWithin the field of prognostics and health management (PHM), health indicators (HI) can be used to aid the production and, e.g. schedule maintenance and avoid failures. However, HI is often engineered to a specific process and typically requires large amounts of historical data for set-up. This is especially a challenge for SMEs, which often lack sufficient resources and knowledge to benefit from PHM. In this paper, we propose ModularHI, a modular approach in the construction of HI for a system without historical data. With ModularHI, the operator chooses which sensor inputs are available, and then ModularHI will compute a baseline model based on data collected during a burn-in state. This baseline model will then be used to detect if the system starts to degrade over time. We test the ModularHI on two open datasets, CMAPSS and N-CMAPSS. Results from the former dataset showcase our system's ability to detect degradation, while results from the latter point to directions for further research within the area. The results show that our novel approach is able to detect system degradation without historical data
Integration of a Skill-based Collaborative Mobile Robot in a Smart Cyber-Physical Environment
Full text linkThe goal of this paper is to investigate the benefits of integrating collaborative robotic manipulators with autonomous mobile platforms for flexible part feeding processes in an Industry 4.0 production facility. The paper presents Little Helper 6 (LH6), consisting of a MiR100, UR5, a Robotiq 3-Finger Gripper and a task level software framework, called Skill Based System (SBS). The preliminary experiments performed with LH6, demonstrate that the capabilities of skill-based programming, 3D QR based calibration, part feeding, mapping and dynamic collision avoidance are successfully executed and strategies for further expansion of the operational capabilities of the system are discussed
Mixed Reality Interface for Improving Mobile Manipulator Teleoperation in Contamination Critical Applications
Full text linkThis paper presents a mixed reality teleoperation interface for mobile manipulation tasks in contamination critical production environments, wherehuman presence is undesirable. This is achieved by using an intuitive control approach and providing the operator with a sense of depth throughvarious visual feedback modalities. The different visual feeds from a mono- and stereoscopic multi-camera setup are displayed for the operator, ina mixed reality control room developed in Unity. The control interface employs the differentiation of the VR controller’s pose, interpolated into atrajectory for the end-effector. The communication between the operator and the robot is facilitated through ROS for control commands and visualfeedback. Speed of operation is typically not crucial in current use cases, while task safety, accuracy, and perception are paramount. The paperpresents the latest research developments of a mixed reality interface designed and tested for a mobile manipulator
Innovation Factory North:An Approach to Make Small and Medium Sized Manufacturing Companies Smarter
No full textThis chapter presents the Innovation Factory North (IFN) as a platform for a collaborative approach to trigger and accelerate industry 4.0 based innovations in small and medium sized manufacturing enterprises. The potentials for SME becoming 'smarter' are huge and well-recognized. However, 'how' to approach this is difficult. In IFN, manufacturing companies, technology vendors, and academia join forces to trigger digital transformation. The IFN approach has been developed during an ongoing regional research and innovation project in a collaboration between industry, academia, and the government. This chapter presents the generalized approach and discuss the preliminary findings from more than 60 cases of applying the approach as steps towards making the participating companies smarter.</p
