Advanced search
Start date
Betweenand

Predictive maintenance of diesel engines for application in agricultural machinery of planting and harvesting - study of technical and scientific viability

Grant number: 17/22755-4
Support type:Research Grants - Innovative Research in Small Business - PIPE
Duration: February 01, 2019 - November 30, 2019
Field of knowledge:Physical Sciences and Mathematics - Computer Science
Principal Investigator:Boris Rotter
Grantee:Boris Rotter
Company:Agriconnected Tecnologia e Inovação Ltda
CNAE: Fabricação de máquinas e equipamentos para a agricultura e pecuária, exceto para irrigação
City: São Paulo
Associated scholarship(s):19/03656-0 - Predictive maintenance of diesel engines for application ín agricultural machinery of planting ánd harvesting - study of technical ánd scientific viability, BP.TT
19/02216-7 - Predictive maintenance of diesel engines for application in agricultural machinery of planting and harvesting: study of technical and scientific viability, BP.TT
19/01483-1 - Predictive maintenance of diesel engines for application in agricultural machinery of planting and harvesting: study of technical and scientific viability, BP.PIPE

Abstract

Predictive maintenance has been increasing the availability and reliability of machines in many industry segments, thus reducing operating costs and enabling safer work environments. However, predictive maintenance is still restricted to electric motors and has not been applied to combustion engines (and in rural environments), due to the inherent particular characteristics of such equipment. With the present research project we propose to investigate the technical and scientific feasibility of the predictive maintenance of diesel engines and also their possible application in agricultural machines. The objective of the project is to optimize maintenance costs and avoid the productivity losses generated by unplanned downtime in the field, especially in agribusiness critical machines such as the ones for planting and harvesting. By installing sensors at the motor, vibration data will be monitored. We will study the possibility of describing the normal and failure behavior of a diesel engine through a fault algorithm and a trend line. The collected vibration data will also be monitored and analyzed by the use of data science and technologies such as Internet of Things (IoT), analytics and machine learning for real-time monitoring and better accuracy and automation in conclusions. The research also involves the development of hardware to be installed on the machine, back-end and front-end software for viewing and sending alerts on critical systems and research on the best form of connectivity to use. The research will focus primarily on engines as these represent the highest maintenance costs for the machine owner or agricultural producer. However, the predictive solution can still be replicated to other equipment. We expect the research to understand and overcome the scientific and technical challenges inherent to combustion engines and to enable a system of data storage and analysis so that we can also develop a statistical and artificial model to evaluate and predict fault behavior using IoT and machine learning. If project feasibility is proven, agricultural machine owners and rural producers will have access to advance information of their machines through a solution that would work regardless of their brand, odel or year of manufacture. The impacts that predictive analysis can bring to the field are many. Precision fault prediction enables the rural producers to better manage maintenance cycles, spare parts, and replace machines that are not suitable for field work, thereby avoiding unplanned downtime and loss of productivity trailer. (AU)