Analytics at the Edge with IOT Data

In an odd way, today’s enterprise challenge is too much data.  Data that is mostly going unused.  A focus on traditional data infrastructure is not sufficient any more, and it is the variety of new technology choices that are to blame. In particular, data at the edge of the enterprise includes the internet of things, IOT, which creates new opportunities and new challenges.

Examples of IOT include automation in warehouses, factories, mass transit, utilities, but also includes getting more out of your video feeds, powering autonomous vehicles, and monitoring jet turbines to optimize service schedules and operational efficiency. The existing IT infrastructure at the edge is usually targeted at solving a specific well-defined problem.  This serves to move processing to where the data resides, and avoids moving data to a central site for processing thereby avoiding network congestion. This is only part of the solution, however.  You’re solving today’s problem.  You have the opportunity to take a step ahead and consider a bigger picture.

Data strategy should play a bigger role in IOT implementations. As an example, the smart factory initiative seeks to automate functions which analyze machine and sensor data at the edge.  Data coming from a specific tool or machine can often be analyzed real-time to improve the operation of the machine.  If there is a data strategy in place, that data can also be used in aggregate to improve the operation of the entire manufacturing process by showing how the machines interact and how solving problems in one area can improve overall productivity.

By using data as a resource, reconsider the problem you are trying to solve. You quickly see that a data strategy can become a valuable tool in deciding how to structure work flows, improve operations, understand your customers and markets better, perhaps even gain a competitive advantage. Data silos exist to solve specific business problems.  One example of how to take advantage of a more comprehensive data strategy would be using public cloud analytic tools to process edge or IOT data and repatriating the results to a central site so it can be used in a broader context for additional benefit.  Think about moving beyond silos to incorporate a broader data strategy to get the full value of your data.  Don’t let IOT data malinger at the edge.  Stomp on the data accelerator to move ahead.

New possibilities in your data strategy emerge as new enabling technologies appear. Fundamental changes in the cost of processing, storage and networking create new ways to monetize your data beyond the IOT silo. Moving compute to the edge has never been easier and less expensive. Reimagine how new technologies can be used for purposes beyond the original intent to bring more value.

There’s been a lot of discussion about how 5G will make mobile and edge data more accessible with significantly higher bandwidth availability and security. It can also be the backbone of private networks and ultimately replace WiFi and perhaps Bluetooth. There are additional technologies that will impact edge and IOT such as NVMe-oF, a low latency highly scalable networking technology that can work over today’s Ethernet with minor adaptation. It’s not just hardware advances either, programs built with containers can make rolling out new applications easier and faster.  More open source APIs allow interaction among different parts of your infrastructure and can improve capabilities at the edge. Composable disaggregated infrastructure is another way to reduce costs and make more efficient use of resources.

Take these infrastructure advances as a way to rethink your data strategy and how to make the most of your data and your business. We are just starting to see the benefits of IOT and edge computing.  Think outside the silo.

For an extreme example of the benefit of analytics at the edge is NASA’s OSIRIS-REx team. They wanted to land a spacecraft on asteroid Bennu and collect a surface sample and return it to Earth. If you look at this short NASA video  you can see how the spacecraft makes a last minute adjustment to the probe to avoid a rock.  This is AI in action, saving a multi-million dollar project at the last minute.  Due to the distances involved, there is no way to control this from Earth, and onboard AI made the save.  You can bet the data from this mission will be reviewed in aggregate for additional insights.

Finding the right mix of compute and capability at the edge for IOT applications and developing a data strategy that makes sure you are getting the full value of your data is a moving target. By keeping the strategic importance of data in mind as you design your IOT solutions you can create a culture of continual improvement and get the full value of your data.