The technology also cut pharmacy costs by $300,000 annually and provided a one-time inventory savings of $400,000.
Challenges to address
Organizations can't expect to reap the benefits of IoT without addressing a number of hurdles. At the minimum there is the need to inventory, bar-code, and cross-check every physical object to be brought online. For Great River, this process, which involved thousands of medications, took several months to complete, Cooley says.
IoT deployments will introduce a host of technical and procedural challenges that companies will need to overcome to reap the benefits of a connected physical network. IoT also involves multiple technology components across the IT stack, so it will require expertise from various parts of the organization -- or perhaps from outside resources.
"IoT by itself is not a technology; you can't buy a box of IoT off the shelf," says Mike Redding, managing director at Accenture Technology Labs, the technology R&D organization within management consulting and outsourcing firm Accenture.
Each technology component brings with it unique issues. At the transponder and reader-device level, this means questions of reliability, battery life, security, access, and data processing.
Network services and application performance is also a concern for IoT deployments, Miles says. For example, a simple sensing and monitoring application for a site with 100 sensors installed and collecting telemetry data might produce raw data totaling more than 4PB in a year.
From a system design perspective, challenges include ensuring effective, data-driven decision making, dealing with a whole new level of data granularity, and determining who owns the data, Miles says. On this last point, for example, under current U.S. health care law patients have a legal right to access their medical records. "But who owns this data, the doctor, the hospital, the software or hardware/services provider?" Miles asks.
Big data like you've never seen before
The biggest hurdle facing organizations considering IoT deployments will be knowing what to do with the massive amounts of information that will be gathered.
"Social media, sensors, and embedded devices expand the ability to gather data from previously unexplored areas," Redding says. "One challenge is to design for analytics -- creating a strategy that sees data more as a supply chain than a warehouse."
As tools mine countless new unstructured data sources, the problem is no longer the absence of enough data, Redding says -- it's making sure you aren't missing out on the data you really need while spending too much on data you don't need.
"With a supply chain of data, organizations can fill the gaps in whatever way is required," Redding says. They can create new APIs to current applications, ask for data from partners or third parties, or create the data by quantifying the business/physical environment around them, he says.
Sign up for Computerworld eNewsletters.