Analytics, On the Edge, and NonStop!

4 Minute Read

It may have gone unnoticed by many viewers of the popular television program, Elementary, when its main character, Sherlock Holmes, heard the buzzing of a mosquito when it wasn’t mosquito season. The buzzing indeed was artificial and we were led to believe that a military subcontractor had reduced an intelligent “thing” performing visual reconnaissance to the size of a simple mosquito. It may have also gone unnoticed by some when more recently, in the television series, Madam Secretary, a SEAL team involved in an extraction mission sent in a slightly bigger insect that was also artificial and yet, carried with it a camera and comms package so that the SEAL team could locate the hostage.


Far- fetched? Preposterous? All quite fictional but somehow, unnerving as we continue to become aware of just how many things today are equipped with sensors and even stand-alone intelligence. Perhaps it is because this is the festive season where gift exchanges are on the minds of many of us, but I was reminded of both these fictional portrayals of new-age intelligence gathering things. However, reading the December 12, 2016, issue of Time Magazine I came across the article, Artificial Intelligence invades the home … in toys.

As the picture above depicts, these playful toys called Cozmo, brought to us by San Francisco startup Anki, have been reduced to the size of a small ball but they are packed with remarkable technology.  “It doesn’t like to stay put very long. Roused from slumber, the little robot’s face illuminates, and it begins zooming around the table in front of me. A moment later, it notices I’m watching and turns to greet me, saying my name with a computerized chirp.”

These toys also differ “In the way they interact with the people and objects around them, changing their behavior over time as their software ‘learns.’ Right out of the box, cameras and sensors allow Cozmo to recognize individuals, avoid falls or bumping into obstacles and play simple games like keep-away.” Furthermore, “‘Every input trigger, no matter what happens to him, will influence his future behavior,’ says Hanns Tappeiner, Anki’s president.”  Suddenly, the Internet of Things (IoT) takes on a whole new dimension – what may start out as a toy has a lot of industrial applications that I know have already started showing up in everything from warehouses to hospitals.

Smart things that can fit into the palm of our hands capable of learning and interacting with us tend to suggest that even smarter devices have already made it into the world’s stage so perhaps what may have been considered preposterous just a year or two ago actually exists. Coming at a time when many of us have taped over the web cam of our laptop, I have to wonder what else may be watching me that I innocently brought into the office. From tradeshows or as gifts awarded for one reason or another; as we accept presents this festive season, I wonder how many of us will be truly scrutinizing everything mechanical as we rip it from out of its wrapping paper.

In a paper focused on his Top 20 Predictions for 2016 that will be published shortly, CTO and Cofounder at Striim, Steve Wilkes, notes how “IoT Platforms will grow in strength and capability incorporating device registration, management and communication features as well as integration, analytics and machine learning.” And that, “Simple IoT use cases such as the real-time tracking of the movement of people or packages via geolocation and time windowing will become prevalent across healthcare, travel, manufacturing and logistics.”

Real-time tracking of the movement of people – well, looks like we already have begun pushing the boundaries of this frontier. IoT will usher us into another era when it comes to how we will buildout our infrastructure – there will be things, there will be centers, and in the middle there will be edges. And for the NonStop community, it’s a given that there will be a NonStop presence in the center but what about the edge? In his paper on predictions, Steve also notes how, in order for timely responses, “Reliability and security concerns will push real-time analytics to edge locations for IoT. This will become evident through connected cars, home IoT hubs, retail store-based gateways and other localized technology. Anonymized data will be pushed to the cloud for deeper analytics.”

In a post to the Striim blog of December 1, 2016, Marketing VP, Katherine Rincon, wrote that, “The edge processing power of the Striim platform running on a gateway based on the Intel® X86 architecture, offers efficiency and expediency to the development and deployment of IoT applications. With the power of the Intel® X86 technology, the Striim agent can seamlessly ingest, processes and analyze data in real time, generating insights in milliseconds.” Furthermore, “IoT applications do not exist in a vacuum. As such, IoT solutions not only need to incorporate a variety of IoT devices, but must also be integrated with the existing enterprise infrastructure.”

When it comes to edge products and the need for them to support analytics capable of operating in real time, such as Striim, the appearance of gateways based on x86 throws open the potential for running NonStop applications on the edge. While much of the arriving sensor information may be discarded following initial analysis, there will be a category of sensor information that simply cannot be overlooked no matter what. A critical message from a pacemaker, an alarm from an overturned vehicle, a mechanical malfunction in an oil refinery or nuclear power station – many events just have to make it into the center and the fault tolerance we have always appreciated from NonStop brings the NonStop system back into play as today, in its new virtual guise as vNonStop, it can run on any x86 server users elect to deploy.