Qualidade de dados e Internet das coisas no framework de rastreabilidade de grãos

Abstract

In the agricultural area, accurate and timely information is needed to enable safer activities in the strategic and operational environment. To ensure the quality of decision making process, the quality of the data must be guaranteed. The use of data with typos, with incomplete information, with null values and others problems, can result in reports that lead to wrong and/or late decision making, causing damage to the organization. Some traceability models allow that users insert manual data about product and process data, which can raise questions about the quality of the system and the data to be traced. Besides, no studies were found in the literature aiming at data quality in agricultural product traceability systems. Therefore, data quality needs to be incorporated into traceability tools to ensure that their processes are reliable. This dissertation aims to specify a grain traceability framework that guarantees the quality of the Attribute to be traced, through a model that considers data obtaining from IoT and data quality criteria. Through the study of agriculturally applied architectures, we specify a layered architecture for the framework. This architecture is evaluated according to parameters of suitability to the Internet of Things. A literature review about the perception of data quality and the identification of quality dimensions in traceability systems define the data quality model for the grain traceability framework. The model obtained is compared with literature studies. As a result, we have the ability to automatically collect data, perform data quality activities for the Tracked Attribute table, define the attribute quality rules, and measure the data quality dimension, using an evaluation method.