Search

Fibrosis characterized by excess accumulation of extracellular matrix (ECM) due to complex cell-ECM interactions plays a pivotal role in pathogenesis. Herein, we employ the pancreatic ductal adenocarcinoma (PDAC) model to investigate dynamic alterations in nanomechanical attributes arising from the cell-ECM interactions to study the fibrosis paradigm. Several segregated studies performed on cellular and ECM components fail to recapitulate their complex collaboration. We utilized collagen and fibronectin, the two most abundant PDAC ECM components, and studied their nanomechanical attributes. We demonstrate alteration in morphology and nanomechanical attributes of collagen with varying thicknesses of collagen gel. Furthermore, by mixing collagen and fibronectin in various stoichiometry, their nanomechanical attributes were observed to vary. To demonstrate the dynamicity and complexity of cell-ECM, we utilized Panc-1 and AsPC-1 cells with or without collagen. We observed that Panc-1 and AsPC-1 cells interact differently with collagen and vice versa, evident from their alteration in nanomechanical properties. Further, using nanomechanics data, we demonstrate that ML-based techniques were able to classify between ECM as well as cell, and cell subtypes in the presence/absence of collagen with higher accuracy. This work demonstrates a promising avenue to explore other ECM components facilitating deeper insights into tumor microenvironment and fibrosis paradigm.

Deep reinforcement learning has produced many success stories in recent years. Some example fields in which these successes have taken place include mathematics, games, health care, and robotics. In this paper, we are especially interested in multi-agent deep reinforcement learning, where multiple agents present in the environment not only learn from their own experiences but also from each other and its applications in multi-robot systems. In many real-world scenarios, one robot might not be enough to complete the given task on its own, and, therefore, we might need to deploy multiple robots who work together towards a common global objective of finishing the task. Although multi-agent deep reinforcement learning and its applications in multi-robot systems are of tremendous significance from theoretical and applied standpoints, the latest survey in this domain dates to 2004 albeit for traditional learning applications as deep reinforcement learning was not invented. We classify the reviewed papers in our survey primarily based on their multi-robot applications. Our survey also discusses a few challenges that the current research in this domain faces and provides a potential list of future applications involving multi-robot systems that can benefit from advances in multi-agent deep reinforcement learning.

Precision agriculture is the collection of hardware and software technologies that allow a farmer to make informed, differentiated decisions regarding agricultural operations such as planting, fertilizing, pest control, and harvesting. In recent years, advances in agricultural machinery and the emergence of agricultural robots continuously increased the resolution at which differentiated treatment is possible. This creates a corresponding need for information at a fine spatial and temporal resolution. Autonomous multi-robot systems (e.g., unmanned ground and aerial vehicles) are some of the most promising approaches for such information collection in open-air farms. In this paper, we survey the current state and challenges of multi-robot information gathering for precision agriculture, with a special focus on maximizing information and ensuring the security of the collected data while simultaneously keeping energy consumption in check.

We consider the NP-hard problem of multirobot informative path planning in the presence of communication constraints, where the objective is to collect higher amounts of information of an ambient phenomenon. We propose a novel approach that uses continuous region partitioning into Voronoi components to efficiently divide an initially unknown environment among the robots based on newly discovered obstacles enabling improved load balancing between robots. Simulation results show that our proposed approach is successful in reducing the initial imbalance of the robots’ allocated free regions while ensuring close-to-reality spatial modeling within a reasonable amount of time.