We’ve talked a lot about data—consumer data, to be specific—and how transforming it into insight could mean either savings, profit, or both for your business. But according to the Computational Ecology and Environmental Science Group (CEES), a part of Microsoft’s Cambridge Research team that we got to meet this month as part of TechFest, ecological data will be just as big of a business opportunity as consumer data. The predictive models these researchers are creating can be applied to many other kinds of business, from Coca-Cola’s manufacturing process to protecting resources.
When physics meets machine learning
CEES combines the knowledge of ecologists and environmental scientists with the skills of software engineers to create predictive models of our environment. Given historical data, the team can create simulation models to show how the natural world responds to various occurrences in the environment.
Data like this can be used to visualize how the distribution of carbon will change in polar versus equatorial regions or to model changes in Earth’s marine biomass over the next decade. Although the association may not be immediately apparent, with insights such as these, major corporations such as Coca-Cola, whose bottling plants depend on local water sources, can act to protect resources. Coca-Cola can use this data to instantly and independently conduct local source vulnerability assessments to inventory risks to the water sources supplying their facilities and the surrounding communities.
Another example would be the construction of data centers: before the first blueprint is ever drawn, enterprises can consider the availability and cost of electric power in addition to the likelihood of disruptive weather events in a particular location. This way, the company will know what to expect in the years to come (e.g., the probability and frequency of server crashes). The same techniques can also be used to understand past events. For instance, to understand why a major server crashed, analysts can use data to reconstruct the circumstances of the event.
When it comes to ecological data helping the bottom line, the challenge isn’t technology—it’s a lack of data. Our data resources are limited because ecologists (and scientists in general) aren’t necessarily the most technology savvy. Whether data is shared in PDFs or stored in an Excel sheet on a researcher’s computer, it remains either locked in a difficult-to-use format or siloed and inaccessible.
Microsoft’s Cambridge Research team is working to compile this data and has already made big strides—from reaching out to scientists and researchers to organizing data from all sorts of public sources. Even major corporations such as Coca-Cola realize what an important asset this data is; as part of its water stewardship effort, the company is “donating data to speed water-risk mapping.” Its goal is to create “an open, transparent and publicly available database that provides geographical and sector-specific water risk context to companies, investors, governments and others.”
Until now, the push to organize ecological data in a structured way has been somewhat of a grassroots effort led (for the most part) by teams such as CEES, nonprofits, and other disparate groups of scientists and researchers. Enterprises are now realizing the potential value of these insights. Beyond the enterprise, how could this insight help your business? Beyond your business, how could it affect your life as a homeowner or human on this planet?