Sensing and Biosensing in the World of Autonomous Machines and Intelligent Systems

Abstract

In this paper we discuss how nanotech-based sensors and biosensors are providing the data for autonomous machines and intelligent systems, using two metaphors to exemplify the convergence between nanotechnology and artificial intelligence (AI). These are related to sensors to mimic the five human senses, and integration of data from varied sources and natures into an intelligent system to manage autonomous services, as in a train station.

Materials Discovery With Machine Learning and Knowledge Discovery

“…This is not so for an intelligent system that relies on a classification algorithm, since as long as the input features and the classes are well defined, the classification task will be performed regardless of the specificities of the data instances. An implication of this principle is the generality of ML algorithms, which may be employed in tasks as diverse as clinical diagnosis with image processing, sample classification with electronic tongues and noses, and even text classification ( Oliveira et al, 2014 ; Rodrigues et al, 2016 ; Paulovich et al, 2018 ; Oliveira and Oliveira, 2021 ; Rodrigues et al, 2021 ).…”

“…In health and in the assessment of athletes’ performance, intelligent monitoring systems will require functional materials for wearable and even implantable biosensors. Similarly, intelligent systems for artificial vision and mimicking the human senses are all based on nanomaterials, as exemplified with electronic tongues, electronic noses, electronic skins ( Paulovich et al, 2018 ; Oliveira and Oliveira, 2021 ; Rodrigues et al, 2021 ). In fact, there is reciprocity in contributions in all of these systems, for research on materials is essential to produce the devices, and ML (or another AI method) is used to analyze the data generated by such devices and make decisions.…”

“…Publishers are already developing AI tools to assist editors, and authors may soon benefit from analogous tools, which makes NLP a key area for science and technology. In fact, a broader perspective of the impact and breadth of AI may be inferred from recent predictions ( Oliveira and Oliveira, 2021 ; Rodrigues et al, 2021 ) that in a few decades the machines themselves may be able to generate knowledge. This new paradigm, with machines capable of generating knowledge without human intervention for the first time in history, will reshape society.…”

Insight into continuous glucose monitoring: from medical basics to commercialized devices

“…Artificial intelligence (AI) including pattern analysis and classification algorithms with biosensors can bridge the gap between the data acquisition and analysis and achieve improved diagnostic and therapeutic accuracy [231]. Today, there is a real need for artificial intelligence that has become a trend in biosensing technology [216,[232][233][234][235]. Indeed, most of the commercial CGMs available now on the market can be connected to insulin pumps to automatize insulin delivery as closed-loop system.…”

Aprendizado de máquina aplicado em dados de biossensores para diagnóstico de câncer e COVID-19

“…(12) De fato, é um grande desafio desenvolver um sistema de diagnóstico que considere dados de análises clínicas, imagens, e informações de saúde e socioeconômicas dos pacientes. (8,13) Destaque-se também a busca por soluções computacionais de maior validade, confiabilidade, capacidade de generalização, e que exibam alto desempenho na classificação de amostras. Das metodologias usadas nesta tese, a mais essencial foi a do Aprendizado de Máquina ( 14), que pode satisfazer os requisitos mencionados acima.…”