Imitation Learning for the Transfer of Human Skills to Industrial Trucks

The increasing demand for efficiency, safety, and cost reduction in logistics in conjunction with rising labor costs is a key challenge. To solve this challenge, sophisticated solutions such as autonomous forklifts, which perform repetitive tasks like material handling without human intervention, are a promising solution. Despite the growing interest in autonomous systems, one of the most significant challenges to overcome is reliable load handling in terms of correctly aligning the vehicle with the load and picking it up correctly. The paper at hand aims to systematically identify, analyze, and synthesize current research activities on this topic and outline current challenges and open research gaps. Various databases were systematically searched for relevant publications, and a total number of 62 publications were analyzed in detail. The study demonstrated that research activities in this field have been gaining importance. Numerous different methodologies and algorithms have been identified and summarized. Nonetheless, challenges and research gaps exist, such as perceptual robustness, environmental adaptability, and system integration. This systematic literature review is the first to address load handling with autonomous forklifts using a systematic and comprehensive approach.

Autonomous forklift, Logistics Automation, Warehousing, Load handling, Mobile Robotics

Industrial trucks play a crucial role in modern logistics, but their effective operation often depends on highly skilled human operators. Although there are already approaches for automated control, these do not have the flexibility, speed of action, comprehension, and experience of human drivers to cope with complex situations. This paper deals with pallet object recognition, which is fundamentally important for the implementation of a pose estimation algorithm. Based on this, this paper lays the foundation for a complete automation solution for industrial trucks.

Artificial intelligence, intralogistics, driverless transport vehicles

This work explores the challenges of fully automating in-house goods transport in environments where industrial trucks like forklift trucks remain necessary due to undefined load carrier positions and shapes. Imitation Learning (IL) is identified as a promising solution for vehicle control in repetitive tasks, yet its application in intralogistics is challenging by the dynamic complexity of industrial trucks and the large dimensional space involved. A Robot Operating System 2 (ROS2) framework is introduced, enabling the acquisition of driving data from both simulation environments and real-world demonstrators. The study also presents a network architecture combining a Convolutional Neural Network (CNN) with a Long Short-Term Memory (LSTM) network, facilitating end-to-end learning from spatial and temporal image data. The framework's effectiveness is evaluated using a dataset of expert driving maneuvers to assess the generalization potential of the IL-trained network in vehicle control in different scenarios. The research aims to demonstrate the utility of the proposed framework for data acquisition and validate IL as a control approach for industrial trucks that require generalization.

Imitation Learning, industrial truck automation, intralogistics, ROS2, load handling

Automated industrial trucks master difficult driving situations worse than humans – for now. New approaches based on artificial intelligence (AI) are intended to replicate human driving behavior and give automated systems more flexibility.

artificial intelligence, intralogistics, industrial trucks