Manufacturing companies are increasingly investing in data collection and analytics to create a competitive advantage.
By Manjunath Kamath
Manufacturing companies are increasingly investing in data collection and analytics to create a competitive advantage. Internal data, such as transactional and log data, has been the primary source of big data for an overwhelming majority of companies, according to a research study conducted by IBM.1
Data visualization and predictive modeling using statistical techniques are some of the core analytics capabilities commonly found in industrial manufacturing. A potential opportunity lies in using the data collected to support an important capability in manufacturing capacity analysis, which could result in long-term benefits to the manufacturing company.
Manufacturing capacity analysis focuses on the ability of system resources (e.g., machines, material handling equipment and operators) to satisfy the demand for goods produced by the system in a timely manner. The amount of work to be completed by the system resources is determined by the volume of demand and the processing requirements of products expressed in terms of product routings, run/set-up times, batch sizes, etc.
Key decisions resulting from capacity analysis include the number and type of resources (equipment and labor) needed given the number of operating hours per day. Typical tools used for such analysis range from load analysis based on averages to probabilistic models such as queueing based on averages and variability to simulation models driven by probability distributions. Load analysis helps in determining minimum resource requirements, queueing models help in identifying resource capacity needed to meet demand under lead time requirements, and simulation models allow for the inclusion of operational details (production/dispatching rules, synchronization, etc.) in fine tuning capacity requirements. With the increasing fidelity in the models and decisions made comes the need for data at an increasing level of detail and granularity.