01 OVERViEW
Automation of Grinding Intelligence (AGI)
-
To extend GPS capability onto the machines itself and to formulate an architecture for it, “Automation of Grinding Process Intelligence (AGI)” project was initiated in industry-academia collaboration with Advance Manufacturing Technology Development Centre (AMTDC) of IIT Madras, funded by Department of Heavy Industries (DHI) of Government of India.
-
AGI brings the evolution of GrindTrak™ getting integrated into a machine for automation of process diagnosis and intelligence.
-
It is a virtual interaction between machine and user with a digital knowledge database of grinding expertise, along with statistical rules and algorithms, automating the signal processing for characterizing the performance of grinding cycle.
02 MY CONTRiBUTION
I was an integral part of the AGI project right from its inception to the completion by spearheading this industry-academia collaboration with my work in various modules that bound an application engineer’s expertise and a signal analyst’s perception. AGI brings the evolution of GrindTrak™ getting integrated into a machine for automation of process diagnosis and intelligence.
I trained my teammate and we together worked on equipping a standard grinding machine with high-speed I/O communication, external 3rd party sensor interface-ability, real-time data acquisition from multiple sensors, visualization of online process as well as remote connectivity.
​
The B&R Industrial PC and Marposs BLU were successfully configured and interfaced with the Fanuc Controller in Master-Slave configuration with PROFINET communication. This was implemented for the first time in India and in Machine Tools Industry, marching in the direction of 4th Industrial Revolution (OR Industry 4.0).
​
Additionally, we also developed a User Interface (GUI) in LabVIEW to demonstrate the visualization of the multi-sensor data simultaneously acquired from the machine and Marposs processor through B&R Industrial PC unit, while being read in a database in SQL Server.
​
Furthermore, it also includes the intelligence soft-side, i.e. the development of a knowledge database, automated signal analysis and process recommendations to the user (An Application Engineer or A Machine Operator).
​
I have provided the industrial support and represented the industrial customer of this project for academia team to cognize the needs of automated processing of signals obtained during the grinding processes and the use of rule-based algorithms for guidelines for problem solving and process improvement. The intent is to bring the science of grinding actively applicable for shop-floor manufacturing.
The intelligent software module went through multiple versions and I have also trained the engineers of Sales and Application departments multiple times in both the plants of MGT for effective usage and necessary feedback.
03 ACHiEVEMENTS
An industry-ready version of this AI-based data-driven software system is being copyrighted with me as one of the inventors by MGTL and AMTDC, IIT Madras.
Standardized Knowledge Database
Brought together all the system inputs and the grinding models (analytical or empirical) into a searchable digital database
In-Process Data Acquisition
Demonstrated real-time data acquisition and logging from multiple sensors, implementing IoT on grinding machines
Automated Signal Analysis
Integrated signal processing and statistical algorithms to give user recommendations for improvement of the process
The work packages that were delivered to achieve the objectives of AGI System has been classified as follows:
​
-
Development of Standardized Data Collection System
-
In-process signal collection demonstrated through AGI machine
-
Analysis and Development Module -
-
Signal Processing and Analysis
-
Development of Rules based on logic, heuristics and process science
-
Implementation and validation of rules
-
Development of dashboard for display of results
-
04 LEARNiNGS
Database
HANDLING
Setting up a database and configuring the cells for specific data types; Accessing the database from third party software applications, such as B&R's Automation Studio and LabVIEW
IOT
CONFIGURATION
Devised an architecture for multi-sensor interface, implemented Master-Slave configuration between three processor/controller units using PROFINET communication
PLC
PROGRAMMING
Programmed the configuration of sensors and the soft PLC - Automation Runtime (B&R), as per the hardware connections in Structured Text language. Also programmed for digital value gauging and SQL communication
Trained Sales and Application Engineers of both the plants of MGT for the use of AGI software and also for implementation of IOT hardware architecture for other machine models as demonstrated in AGI machine