These past few weeks I have been embroiled in numerous conversations about the direction being taken by automation. Whether it's a factory production line, a farm or even related to planes and cars, automation seems to be at the fore of most predictions about the likely future for us all. No longer tied to doing anything by rote, in theory we should be capable of free thinking and coming up with new ideas as leading vendors, including HPE, view ideas as the driver of future business growth.
It was at HPE Discover 2016 that Meg Whitman told the audience of HPE's biggest event of the year how, "Transformation is fueling the idea economy … driving better business outcomes." Collectively, we all can see the outcome from undertaking the type of transformation (to hybrid infrastructures) that HPE promotes as involving considerably greater automation and HPE is among the major vendors investing in key infrastructure components needed by automation – but the inherent complexity that has arisen tends to require a lot more than just automation. It needs a lot of smart people – smart, and experienced. This is a topic I have touched on in former posts but the phrase augmented intelligence (and not just artificial intelligence) is making waves in IT and will likely find supporters in the data center.
"In the past two years, two highly automated and properly functioning airlines crashed, the Air France plane in the Atlantic and the Colgan Air plane in Buffalo, New York," was posted to the August, 2016, issue of the automotive magazine, Motor Trend. "When automation entered the cockpit, accident rates declined, as experienced pilots were aided by sensors, warnings, and eventually automated corrective features built into the controls." Almost all of us have reason to fly these days and as we board our flights, we often cast a quick glance at the cockpit – and it's almost exclusively a collection of flat screens. In one sense, it's encouraging to see the deployment of new technology and yet, sometimes we are left wondering. Do today's pilots really have all the skills to intervene should circumstances arise that call for exceptional flying skills?
It's an unfortunate reality, the above post noted, that "today's airline pilots transition from hand-flown planes to highly automated aircraft early in their flying careers. And as we see from time to time, they lack some basic flying skills as their roles changed from pilots to systems monitors, with resultant accidents that wouldn't have occurs years ago to any pilot with average flying skills." The key message here can be found right at the outset of this post, "accident rates declined, as experienced pilots were aided" and when it comes to today's complex systems, we are possibly at that stage where application performance monitoring solutions are beginning to cross that line and provide their own "automated corrective features."
In a few days' time I will be participating in the first big kick-off event of the financial year for IR. I look forward to these events each year and they have now become a regular feature on my calendar. When it comes to providing application performance monitoring solutions, IR has both the experience and ability to deliver what has been established over the years as the best in class product offering for the NonStop community – something that is hard to miss, and prominently stated on IR's web site's home page. More importantly for me, each big kick-off event gives me a lot of material that eventually finds its way into posts to this IR web publication, and automation and indeed artificial intelligence have gained more attention of late simply because they hold the promise of ensuring the maximum level of uptime – even for NonStop systems!
After last year's event, two posts proved popular. The first, posted July 29, 2015, More NonStop transactions? More for Prognosis to monitor! makes reference to opening remarks by IR CEO, Darc Rasmussen. "The voice of the customer is the most important influence," said Rasmussen. "It only matters if it makes a difference for our customers! Without customers, you have no business. And business is engaged in transactions and without transactions our society wouldn't work!" An important point to touch on as it is the world of transactions that drive the requirement for greater management oversight of all passes through today's complex mesh of systems and networks.
However, having customers and being involved in transactions is just the starting point. And this too was made very clear at last year's event. The second post of August 28, 2015 For NonStop users the projected upward path of Prognosis heralds good news addressed where application performance monitoring solutions were headed. A trajectory was defined where the first two points plotted on the trajectory were well known – visibility and insight. Seeing what was happening and then being provided with the insight to respond. However, the next two points plotted along the trajectory provided further insight as to where IR was taking Prognosis. Prediction and prescription followed by self-healing models heralded a transition to embracing big data and artificial intelligence.
And this is where the storyline gets interesting. What makes the prospect of turning to artificial intelligence fed by big data together with real time data stream analytics is that the model for an operational "auto pilot" that evolves is developed by IR following years of real world experience even as it aids the decision-making of those who are ultimately responsible for the NonStop systems. All systems have limits and what IR is clearly setting out to do is to equip those in charge with all that they need to ensure that maximum level of uptime for the entire mesh of systems and networks is achieved.
In the July 1, 2016, issue of Fortune Magazine appears the short article, "Knowing the limits of machines." Fortune columnist Geoff Colvin concludes the article with the important observation, "Technology may reduce the number of people in such roles – it's already taking over tasks of middle managers, for example – but responsibility ultimately ends up in human hands." He then closes with "Just because a technology can do a job brilliantly doesn't mean that it should." Pilots, as has already been noted, need first to learn how to fly and fly well before the flight augmentation on hand can prove useful.
Same with cars; as we read recently following a Tesla car's crash Tesla assured us its car turns off its autopilot mode of operating as soon as it detects that the human's hands are off the wheel, and that exactly what happened seconds before the car crashed (read more in the LA Times)! Likewise, and it all seems rather obvious now, along with flying planes and driving cars, every data center operator needs to learn everything about the applications, systems and networks in their charge before the best operations augmentation can prove useful.
I anticipate hearing much more on this topic in the coming days even as I know there will be considerably new information for me to digest before inclusion in future posts. Check back shortly to find out, but even as automation is in full-creep mode with how much real driving are we doing these days I am reluctant to suggest we can turn everything over to the programming models. What truly encourages me when it comes to IR and Prognosis is that where NonStop is involved, operations augmentation is already emerging and for that, the uptime of NonStop being maximized is a reality!
