Supply chain analytics

Supply chain analytics

1. Techiques

   • Statistical analysis
     •  deal with the supply chain uncertainty  
     • inferential
       •  deduce properties and make predictions about a population  Sunil Tiwari (1)  short- or intermediate-range decisions  Gang Wang (2)   
       • qualitative
         • no past data
           •  subjective judgment of consumers or experts  Gang Wang (2)   
       • quantitative
         • time series analysis/forecast
           •  predicts the future  Gang Wang (2)  
         • regression analysis
           •  understanding relationships and causality effects between variables  Gang Wang (2)  
     • Descriptive statistics
       •  describe or summarize the feature of sample or population   Sunil Tiwari (1) 
       • numerical calculations
       • graphs
       • table
   • simulation
     • allow the diagnostic, predictive/ prescriptive analytics   
     • run the “what-if” analysis under different system configuration and complexity
   • Optimization
     • draw system with factors+ constraints
     • uncover new data+analyze multiple objective

2. data categories

   • Descriptive
     • data
       • current/previous happening process
     • examples
       • Standard reporting, ad-hoc reporting, dashboards, querying, drilling down
     • why
       • identify problems and oppotunities
     • researches
       • most approach: association, apply for all stages in SC
       • rarely focus: visualization model
   • Predictive
     • data
       • real time/ batch time processes
     • examples
       • root-cause analysis, Monte Carlo simulations, and data mining
     • why
       • accurately predict the future ( demands/sale..)
     • researches
       • most popular: classification (production planning/control/maintenance/logistics/procurement)
         • truong nguyen (3)
       • semantic analysis ( i.e fraud detection/behavior-based safety analysis)
         •  Truong Nguyen ( 3)  
   • Prescriptive
     • most focus ( Truong Nguyen 3)
     • how
       • data +mathematical algorithms
     • examples
       • hybrid data/ integrated predictions and prescriptions/prescriptions and side effects/ adaptive algorithm/ feedback mechanism
     • why
       • determine and access alternative decision
     • researches
       • most popular approach: optimization ( logistics/transportation)
         • truong nguyen (3)

3. future researches

   • application to SC function level
     • quality control in manufacturing
     • dynamic vehicle routing in-transit inventory management in logistics/transportation
     • demand shaping in SC
     • order picking and inventory control systems in warehousin
     • operational research/procurement
   • horizontal integration of BDA-driven SC
     • vehicle routing and facility location,
     • supplier selection and order allocation
     • demand-driven storage assignment and order picking.
   • Three levels of analytics should be equally examined
   • Combine different data analytic techniques
   • Application of BDA on closed-loop SCM
     • manufactured/remanufactured items are pushed forward to the markets, and backward for recovery once they have reached their useful life. 
   • New tools and BDA techniques
     • reality of distributed systems