This article studies how to improve the generalization performance and learning speed of the navigation agents trained with deep reinforcement learning (DRL). Although DRL exhibits huge potential in robot mapless navigation, DRL agents performing well in training scenarios are often found to perform poorly in unfamiliar scenarios. In this work, we propose that the representation of LiDAR readings is a key factor behind the degradation of agents' performance and present a powerful input preprocessing (IP) approach to address this issue. As this approach uses adaptively parametric reciprocal functions to preprocess LiDAR readings, we refer to this approach as IPAPRec and its normalized version as IPAPRecN. IPAPRec/IPAPRecN can highlight important short-distance values and compress the range of less-important long-distance values in laser scans, which well address the issues induced by conventional representations of laser scans. Their high performance was validated by extensive simulation and real-world experiments. The results show that our methods can substantially improve navigation agents' generalization performance and greatly reduce the training time compared to conventional methods.

Heat accumulation is a critical problem in continuous multi-layer laser aided additive manufacturing (LAAM) process, resulting in inhomogeneous mechanical properties and non-uniformity in the deposited height which can deteriorate the deposition process. This work presents a new integrated finite element (FE) simulation and machine learning approach to select a multi-layer laser infill toolpath planning strategy for fabricating quadrilateral parts to minimise localised heat accumulation during the deposition process. After one layer deposition simulation, the approach employs a Temperature-Pattern Recurrent Neural Networks (TP-RNN) model to predict the temperature field after the next layer deposition for each of the candidate infill toolpaths, and a process parameters inspired thermal field evaluation method to select the best candidate toolpath. The approach would significantly improve the computational efficiency of the laser infill toolpath planning, which was validated by improving the flatness of the 20-layer cube deposition samples with two dimensions (20 mm × 20 mm and 30 mm × 30 mm).

Developing physiologically relevant in vitro models for studying periodontitis is crucial for understanding its pathogenesis and developing effective therapeutic strategies. In this study, we aimed to integrate the spheroid culture of periodontal ligament stem cells (PDLSCs) within a spheroid-on-chip microfluidic perfusion platform and to investigate the influence of interstitial fluid flow on morphogenesis, cellular viability, and osteogenic differentiation of PDLSC spheroids. PDLSC spheroids were seeded onto the spheroid-on-chip microfluidic device and cultured under static and flow conditions. Computational analysis demonstrated the translation of fluid flow rates of 1.2 μl min⁻¹ (low-flow) and 7.2 μl min⁻¹ (high-flow) to maximum fluid shear stress of 59 μPa and 360 μPa for low and high-flow conditions, respectively. The spheroid-on-chip microfluidic perfusion platform allowed for modulation of flow conditions leading to larger PDLSC spheroids with improved cellular viability under flow compared to static conditions. Modulation of fluid flow enhanced the osteodifferentiation potential of PDLSC spheroids, demonstrated by significantly enhanced alizarin red staining and alkaline phosphatase expression. Additionally, flow conditions, especially high-flow conditions, exhibited extensive calcium staining across both peripheral and central regions of the spheroids, in contrast to the predominantly peripheral staining observed under static conditions. These findings highlight the importance of fluid flow in shaping the morphological and functional properties of PDLSC spheroids. This work paves the way for future investigations exploring the interactions between PDLSC spheroids, microbial pathogens, and biomaterials within a controlled fluidic environment, offering insights for the development of innovative periodontal therapies, tissue engineering strategies, and regenerative approaches.