Road map for implementing AI-driven Oulu Smart excavator

Abstract

<div lang="en" class="abs"><p class="abs_header">Abstract</p><p>Evolving machine control systems for excavators are getting more capable every year in civil engineering, now they are usually equipped with hydraulic motion control, localization, and design models in form of building information modelling (BIM). Machine control systems are advancing side by side with the adoption of fast wireless connections like 5G and growing trends of internet of things (IoT) and machine learning. AI across our ecosystem has made autonomous excavator more ubiquitous in nature. The autonomous excavators have gained significant interest in earth works area, due to their enhanced productivity for long hours, safety and lack of skilled human operators, and space exploration for unmanned mining and construction work.</p><p>However, A great amount of effort is required to address many existing challenges such as, adaptive movement and control, task planning (digging, moving debris etc.), continuous environment sensing, avoiding collision (moving animals or objects on site), collaborative work with other machines and humans. In this study, we review state of the art and provide a artificial intelligence (AI-) driven road map for implementing a complete autonomous framework for earthmoving machine to our autonomous excavator test platform ’Smart Excavator’. Furthermore, the challenges and required effort to implement the framework are also discussed in comparison with existing literature.</p></div>