A mining operation is dependent on solid supply lines. From MSHA (Miner's Safety and Health Administration) regulations to profit margins, you might say a supply chain keeps the whole thing together. In a coal mining operation, product is excavated from the ground then sent down a conveyor belt to a silo under which passes a freight train that's incrementally loaded with product until it has been filled. This product is then shipped across the country. It is all done very exactly with precisely defined time-tables and scheduling; or rather, that's how it should be done.
In Gillette Wyoming, there are two mines run by the same company: Eagle Butte and Belle Ayr. Hypothetically, one mine could be in a delay while the other functions ahead of schedule. Is there any reason to let an empty train remain dormant at a mine experiencing a delay when the company's other production site has a surplus? Yet without real-time mathematical optimization, this is exactly what could happen. The very nature of supply chain breeds unpredictability, and having solid metrics continuously measuring supply data is the best way to maximize utilization of resources and avoid unprofitable inactivity.
Mining and metal companies aren't the only ones with such problems — it's inherent across any process manufacturing company (oil and gas, pulp and paper, consumer packaged goods and so on). Sudden changes can trigger increased / decreased demand, supply shortages, inventory glitches and How To Interpret The Information Correctly
Big Data has been the "it" word for several years now. However, companies are now facing another challenge: what do they do with all that data? That's where algorithms and mathematical optimization come into play. The best way to process all the data companies now have access to is through software which uses algorithms to mathematically optimize operations. In fact, it's the only way to properly utilize data for accurate business planning insight.
Proper mathematical optimization can predict a negative trend before it wreaks havoc on those managing operations. A great example is in the aforementioned coal mine scenario. Say demolition experts are about to blast a great swathe of overburden (that's: the dirt above "product"; in this case coal), and someone accidentally miscalculates the proper amount of explosive, sending chunks of rock across the path of excavation equipment like CAT haul trucks half the size of an apartment block. The entire mine has to shut down while bulldozers clear the wreckage from the coal road.
Without algorithms that enable mathematical optimization, it's hard to tell whether or not a train should still route to that mine, or head to the company's other location. But with optimization, those variables can be fed into the software and algorithms will denote the best course of action.
Mathematical optimization obviously doesn't only apply to companies in process manufacturing. However, to stress the monetary value of algorithms and mathematical optimization in business, we’ll stick with a real-world oil and gas problem that could potential cost thousands — or millions — without it.
A fully-loaded Unit Rig haul truck is about 1.2 million pounds; half of which is product. That's about 300 tons (though the big CAT-trucks carry 650). Typically, in Wyoming mines a coal truck can make anywhere between 12 and 30 "runs" daily. That's between 3,600 and 9,000 tons per truck, per day. If there are twelve trucks running 24-hours a day, that's between 86,400 and 108,000 tons of product. According to Quandl, a "short ton" (2,000 lbs) of coal is worth $43.50 in today's market. That means a coal mine operating at the rates defined in this paragraph makes between $4.01 and $4.7 million dollars a day. If there's a six-hour delay because of a blasting operation gone wrong, that's a million dollars lost for the shift. Should optimization software be able to predict such a problem as it crops up and re-route accordingly, quotas are preserved, loss is curtailed and expansion is more likely.
Leveraging algorithms for mathematical optimization in today's market space allows companies to apply greater computational horsepower to operational endeavors than ever before. Because they’re able to process data faster, they’re able to anticipate and avoid potentially disastrous, costly mistakes. Algorithms make it possible for companies to actually use all that data in order to increase profit margins quantitatively. With this in mind, it’s easy to understand why Gartner has coined “algorithmic business” and mathematical optimization as the future of business success. After all, “it’s not just about big data; It’s what you do with it.”