Our previous post on the Internet of Things cited the approaching wave of Big Data—and the demand analytics that make sense of all the bits washing ashore. Here are five essential takeaways:
1. More data and advanced analytics are changing the game in the supply chain—promising an increase in forecast accuracy over traditional statistical algorithms.
Panda agrees, suggesting it’s no longer about bringing divergent viewpoints together into one version of the truth to drive Sales and Operations Planning (S&OP). He says that big data technologies, sensors and telemetry offer a rich source of potential insights.
Costa Express, highlighted in a previous blog, is a good example. At Costa, machine telemetry feeds real-time data for demand sensing and replenishment from 3000 self-serve coffee locations. By the way, Chris Clowes, supply chain manager at Costa, will be a featured presenter at Lora Cecere's upcoming Race for Supply Chain 2020 web seminar.
3. Sensing data yields a more responsive supply chain, particularly with volatile demand.
Panda says, “RFID chips, point-of-sale systems and beacon technologies generate volumes of information that were once difficult to analyze.” Demand analytics can now do that heavy lifting. He adds that, the two-way flow of information via the Internet of Things is particularly crucial in “enterprise operations where production may need to match highly variable demand patterns.”
4. Sensing data also reduces the bullwhip effect.
h and larger swings in inventory in response to changes in customer demand—is caused by belated response to demand. Analyzing Big Data “extends the ability to respond, to predict and, in some cases, even recommend subsequent action,” Panda says. Supply-chain strategy is evolving from “‘supply to forecast’ to ‘supply to prediction based on dynamic pattern analysis.'"
5. Machine learning is key to adaptive supply chain planning.
Cognitive computing is “learning” computing, well-suited to the complexities and ambiguities of adaptive planning. Combined with an expert planner’s “insights of how to respond to specific situations, these technological solutions could enhance the overall responsiveness of the business,” Panda says. Cognitive computing is key “where there are numerous variables to be considered simultaneously.”
Next week we’ll explore a new approach to demand forecasting called “Demand Modeling”.
To better understand how to take advantage of the explosion of already available data in your forecasting environment, check out the infographic and whitepaper below.