“Just because you have big spreadsheets doesn’t mean you have Big Data,” Hayward explains. “It’s not just a dataset that has a thousand data points.” In his role at AIHA, Hayward has worked closely with the Content Portfolio Advisory Group, which identified Big Data as one of AIHA’s research priorities a few years ago. Hayward organized a meeting on Big Data at AIHA’s Falls Church, Va. headquarters in November 2017. The participants at that meeting included representatives from CPAG, the AIHA Computer Applications Committee, the International Safety Equipment Association, NIOSH, software vendors, and technology companies, including Google. The goal of the meeting was to identify opportunities for AIHA to play a leading role in the application of Big Data to industrial hygiene. As was discussed at the meeting, Big Data has several characteristics that differentiate it from regular data. The most important characteristics are known collectively as the “three Vs”: volume, variety, and velocity. Volume is the amount of data being generated in real time; typically, Big Data is measured in terabytes (one terabyte equals 1,000 gigabytes). Variety refers to the types of data; Big Data usually entails many types of data, including images and videos in addition to text and numbers. Velocity is the speed with which the data is generated and processed. Big Data analysis typically involves the collection of vast amounts of data from many sources in real time or near-real time, which is processed and output in a matter of seconds according to predetermined algorithms.  The literature on Big Data identifies at least four other Vs (veracity, value, variability, and viability), but the participants at the AIHA meeting agreed that the three-V model is sufficient for understanding how Big Data might apply to industrial hygiene. Do these characteristics exist for the kinds of data most industrial hygienists have access to today? Hayward doesn’t think so, but he stresses that it won’t be long before they do. “We’re on the cusp of having variety,” Hayward says. “We have some instruments that are able to take data and correlate it to GPS coordinates, which are then manually put on maps or process diagrams. And for years, we’ve been putting noise measurements on unit plot plans. What we haven’t been able to do is combine those processes into one step and then analyze dozens, hundreds, or thousands of data points to form exposure estimates. Right now a lot is still manual, but the instruments we have today are continuing to evolve, with both volume and velocity increasing. Variety still needs to be further developed.” Hayward adds that concerns around the privacy and security of sensitive information need to be a top priority in this discussion before Big Data analysis could be applied across the board to industrial hygiene datasets. Some companies consider exposure data to be medical or personal information. In the U.S., the Health Insurance Portability and Accountability Act of 1996 requires the safeguarding of medical information. Privacy laws in Europe are even more strict. In addition, individual companies likely have internal policies that govern the sharing of personal information and data. Security and privacy need to be factored into any attempt to use Big Data analysis, Hayward says. Big Data originally grew out of the efforts of internet companies such as Google, Amazon, and Facebook to monetize vast amounts of data about individuals’ online behaviors. For example, consider someone who needs new shoes. He searches online for local shoe stores, and then he finds an ad for Converse Chuck Taylors in his Facebook feed. Clicking on the ad opens a page on Amazon’s website where he can purchase the shoes. Meanwhile, he is served suggestions for products that other Converse customers bought on Amazon. This is Big Data as marketing: automatic, algorithm-based delivery of the “right” ads at the right time, determined by voluminous data about the online behaviors of similar customers. In recent years, as many kinds of data, including data collected by the federal government, have become available online, the computing tools for discovering signals in the noise have been refined. The same or similar tools that the internet companies use to parse online behaviors are available in “the cloud”—a network of servers hosted on the internet—and can be used for other purposes.  Other companies see potential in harnessing those tools for worker protection. One such company, Corvex, has created an open-license software platform that connects workers’ smart devices to a system of beacons that tracks their location as they move through a facility. According to Corvex’s founder, Joe O’Brien, who attended AIHA’s meeting on Big Data, the Corvex platform is being piloted at a million-square-foot manufacturing facility in Minneapolis. The facility is divided into 32 zones, some with specific requirements for entry. As workers approach a zone, the platform delivers information to their smart device about the safety plan for that zone, as well as the training and personal protective equipment required to be present in that zone. Workers can also take pictures of hazards—a spill, for example—and submit them to the platform, where they join a queue for resolution. The goal is to encourage workers to participate in their own health and safety. “While this platform is being used, it’s collecting and distributing massive amounts of relative data,” O’Brien says. “It alerts the entire work force to potential risks and ongoing warnings. And it allows for a predictive element that can help improve policies and processes.”