A single Downtime in manufacturing process impacts entire production factors.The reasons that could impact can be anticipated by exploring and connecting real time dependencies.The butterfly effects that can impact the production resources are to be identified for uninterrupted process.There is a lot of real-time data of various factors plant rely on like environmental,electricity/power,social,human resource factors etc.Finding and Linking that data to plant data helps to predict manufacturing outages which saves resources and improve plant efficiency.
What it does
Obtain real time data relating to the plant data and warn if any possibility of impact to process.It is done bottom to top approach by fixing the production events and finding factors that may effect i.e., by observing causality with location dependency,weather dependency,power dependency,social dependency.If the event is sensible to any of the factors the event is attached to real-time data of that factor.ex:Logistics for raw material acquiring or shipping to warehouse depends on weather,traffic,social factors so real-time data of the route is correlated with fleet data and warn if any possibility of delay.Similarly for machine dependent process power outages are monitored.
How I built it
Obtaining real-time data of factors identified from bing maps for location,traffic,route data,bing news to get social regional factors,openweathermap for weather data,eia and national grid for power cut data.They are synchronized with plant data.If any factor is active the app warns to reschedule or to plan for impact.The entire data is formed as Bayesian network and causalities are identified and probability of process is notified.
Challenges I ran into
There is no open data of plant process so based on data from various resources the functioning is assumed.how to relate it with GE api's is main place of difficulty.Finally using ROUTE service created route which covers all process flow from acquiring raw material to storage at warehouse
Accomplishments that I'm proud of
Utilizing and correlating all real-time data ,get Acquaintance with GE api,Applying Bayesian network to real time scenario.
What I learned
working with Bing apis,GE apis
What's next for NICIRA
The data and networks formed are saved for deriving insights and analytics for improving performance and effeciency of plant.Alert or Notify via message/mail