Particle counting serves as a pivotal constituent in diverse analytical domains, encompassing a broad spectrum of entities, ranging from blood cells and bacteria to viruses, droplets, bubbles, wear debris, and magnetic beads. Recent epochs have witnessed remarkable progressions in microfluidic chip technology, culminating in the proliferation and maturation of microfluidic chip-based particle counting methodologies. This paper undertakes a taxonomical elucidation of microfluidic chip-based particle counters based on the physical parameters they detect. These particle counters are classified into three categories: optical-based counters, electrical-based particle counters, and other counters. Within each category, subcategories are established to consider structural differences. Each type of counter is described not only in terms of its working principle but also the methods employed to enhance sensitivity and throughput. Additionally, an analysis of future trends related to each counter type is provided.

This study presents an innovative material classification system designed using an inductive tactile sensor and machine learning algorithms. A simple-structured sensor based on the principle of electromagnetic induction was developed to capture varying inductance signals induced by different materials with distinct magnetic properties, facilitating material detection and distinction. A dataset comprising 10 types of materials was evaluated with the sensor, and three machine learning algorithms, namely the support vector machine, k-nearest neighbors, and naïve bayes models, were trained using the output data. Subsequent performance evaluation employed several metrics, including mean accuracy, precision, recall, and others, and revealed that the naïve bayes model exhibited superior performance in prediction. Finally, an enhanced aggregated classification model was developed, where the soft voting ensemble learning technique was employed with the individual classifiers mentioned above as base models. The study underscores the system’s feasibility for potential implementation in high-performance manufacturing and intelligent automation, such as the motorsports and automotive sector, which could facilitate the development of an Industry 4.0 environment. Furthermore, the study also suggests routes for future work that could bolster performance of this system and emphasizes on the necessity for practical implementations to link the system with real-world applications.