Search String of the Day (1) - ROBOTICS

Coming from an oil & gas background, I like to keep up with technologies that can help us transition to cleaner sources of energy. A key word here is “transition.” While we are moving toward renewables, we will still be using petroleum in some form for some time to come.

So, the technologies that interest me include those that minimize the impact of oil & gas on the environment.

Which brings me to the subject of this post … Google search strings.

The featured search string of the day is …

oil gas robotics UAV

Robotics and UAVs – Unmanned Aerial Vehicles – are an important component of the inspection of pipelines and oil spills, to pick a couple of things.

Three articles from a recent search, all available in open access PDFs, appear below.

But a key take away from this post is that you can use this simple search as a framework for your own purposes, focusing on your own industry. Simply substitute, for example, “wind” or “wind farm” for “oil gas,” and Bob’s your uncle.

///////
THREE RESULTS FROM A RECENT SEARCH ON oil gas robotics UAV

[PDF] Boundary tracking of oil spill and acquittal life protection in seawater using swarm robotics
A Gupta, K Daiya, R Jain - network, 2023
Abstract— This study takes a unique solution to the critical environmental issue of oil spills in coastal areas, combining materials science, robotics engineering, and wireless communication technologies. Swarm robots are used to track oil spill limits while protecting marine life in saltwater conditions. Traditional oil spill response strategies have shortcomings, such as delayed discovery and inadequate marine environment protection. Our multidisciplinary method combines buoyant robots with greater navigational stability using materials such as PLA and foam. These robots are outfitted with sensordropping mechanisms for real-time oil spill detection and wireless communication to improve cooperation. An audio warning system keeps marine creatures away from potentially dangerous places. Through simulations and controlled experiments, the study confirms the system's functionality, exhibiting precise boundary tracking and effective oil spill detection. This study provides a proactive method for mitigating the environmental repercussions of oil spills and has the potential to benefit both society and industry.
Source: https://www.kartikeyadaiya.com/wp-content/uploads/2023/10/RESEARCH-SWARM-ROBOTICS-PROJECT-1.pdf

[PDF] Autonomous Advanced Aerial Mobility--An End-to-end Autonomy Framework for UAVs and Beyond
S Mishra, P Palanisamy - arXiv preprint arXiv:2311.04472, 2023

[Jean’s note: I particularly like this article for its clear description of the various elements of autonomous systems and AI-Artificial Intelligence. As a layman, I can find the whole AI thing very confusing. The excerpt below helps me navigate the morass of terms and concepts involved.]

1) Rule-based vs Learning-based Approach
Autonomous systems have learning-based methods at their core. Artificial neural networks (ANN) are universal function approximators; that is, it is possible to represent complex nonlinear behavior in a high-dimensional space using ANNs. A deep neural network is an ANN with multiple hidden layers and nodes cascaded between input and output layers. Deep neural networks are sophisticated neural networks that have been successfully applied to analyze data in many disciplines in the past several years such as computer vision, image recognition, automatic speech recognition, bioinformatics, finance, and natural language processing [19]. In general, traditional machine learning algorithms such as decision trees, Naïve Bayes classifiers, K Nearest Neighbors etc. are particularly task specific. However, deep learning networks are capable of learning intricate structures in large datasets, allowing them to generalize better to address all the scenarios – however non-linearly related – that are included in the training dataset. Additionally, deep learning algorithms typically do not require the type of extensive feature engineering that is required of other traditional machine learning methods4.
Another class of learning-based approaches that have the potential to enable full autonomy is Deep Reinforcement Learning (DRL). DRL is a branch of machine learning that combines deep neural networks with reinforcement learning, a technique that learns from trial and error by interacting with an environment. DRL agents can learn complex and optimal policies for sequential decision-making problems, such as controlling an aerial vehicle, without requiring explicit supervision or prior knowledge. DRL has been successfully applied to various domains, such as robotics, games, and self-driving cars [20] [21] [22].
Rule-based approaches, on the other hand, do not have any generalization capabilities. Furthermore, closed form analytical equations based models do not account for the changes in the environment and rigidly follow the constructs with which they are written by humans. There is no “continuous” improvement process that allows rule-based approaches to evolve to cater to higher and higher machine intelligence needs. For applications where the number of scenarios and variations are vast – it is extremely challenging, inefficient, and practically close to impossible to create intelligent systems using rule-based programs that can sense, perceive, understand, and act in real-time for dynamically changing environments.
[ page 4 ]
Source: https://arxiv.org/ftp/arxiv/papers/2311/2311.04472.pdf

[PDF] Versatile Airborne Ultrasonic NDT Technologies via Active Omni-Sliding with Over-Actuated Aerial Vehicles
T Hui, F Braun, N Scheidt, M Fehr, M Fumagalli - arXiv preprint arXiv:2311.04662, 2023
Abstract. This paper presents the utilization of advanced methodologies in aerial manipulation to address meaningful industrial applications and develop versatile ultrasonic Non-Destructive Testing (NDT) technologies with aerial robots. The primary objectives of this work are to enable multi-point measurements through sliding without re-approaching the work surface, and facilitate the representation of material thickness with B and C scans via dynamic scanning in arbitrary directions (i.e. omnidirections). To accomplish these objectives, a payload that can slide in omnidirections (here we call the omni-sliding payload) is designed for an over-actuated aerial vehicle, ensuring truly omnidirectional sliding mobility while exerting consistent forces in contact with a flat work surface. The omni-sliding payload is equipped with an omniwheel-based active end-effector and an Electro Magnetic Acoustic Transducer (EMAT). Furthermore, to ensure successful development of the designed payload and integration with the aerial vehicle, a comprehensive studying on contact conditions and system dynamics during active sliding is presented, and the derived system constraints are later used as guidelines for the hardware development and control setting. The proposed methods are validated through experiments, encompassing both the wall-sliding task and dynamic scanning for Ultrasonic Testing (UT), employing the aerial platform - Voliro T.
Source: https://arxiv.org/pdf/2311.04662.pdf
///////
Google® Better!
Jean Steinhardt served as Librarian, Aramco Americas (https://americas.aramco.com/ ), Engineering Division, for 13 years. He now heads Jean Steinhardt Consulting LLC, producing the same high quality research that he performed for Aramco.

Follow Jean’s blog at: http://desulf.blogspot.com/ for continuing tips on effective online research
Email Jean at jstoneheart@gmail.com with questions on research, training, or anything else