Page 27 - AMTDC Annual Report 2019-20flipbook
P. 27

When a new grinding process is to be designed, process
          engineers refer to handbooks, vendor recommendations,
          prior experience and other in-house expertise. But even
          with all these sources, it is not very easy for a process
          engineer to integrate all this information and determine
          the optimal parameters. The first part of AGI project is to
          develop  a  document/database  containing  all  the
          information related to grinding process.

          The second phase of the project is to assist experts (and      Process Intelligence into
          even  the  non-experts)  in  finding  the  relationship
          between the input and output parameters of the process         Industry 4.0
          as available from the system document and diagnose any
          issue  or  further  optimize  the  process.  As  mentioned     AGI  is  the  only  intelligent  software
          earlier, grinding is a complex process involving multiple      platform which  integrates  the  process
          process parameters. And it reflects a set of well-defined      knowledge  related  to  mechanics  and
          microscopic interactions and their stability or change. To     microscopic  interactions  with  Data
          understand  the  microscopic  interactions  and  find  the     Science  compared  to  the  other
          relationship between input and output parameters, one
          has  to  record  the  microscopic  interactions.  The  in-     platforms  which  offers  only  simple
          process signals collected during the process will be used      industrial solutions.
          to record and analyse these microscopic interactions.          Even  the  non-experts  of  grinding
          The third phase of the project is to automate the above        process will be able to use the AGI for
          process  of  problem  identification  from  the  in-process    problem  solving,  optimization  and
          signal, its analysis, rules and system document. For this      troubleshooting the grinding process.
          the in-process signals obtained has to be processed and
          converted into quantifiable signal features using various
          signal processing and feature recognition algorithms.

          The fourth and final phase of the AGI project is to develop
          a dashboard. This dashboard which can be installed in any
          grinding machine. Ultimately the results will be subjected
          to  analytics  and  data  science  to  identify  further
          inferences  about  the  manufacturing  process  (unique
          asset)  by  itself,  in  combination  with  other  similar
          operations within a plant or within a company, etc.


                                                           26
     AMTDC Annual Report | 2019-20
   22   23   24   25   26   27   28   29   30   31   32