Here at SevOne, we spend a lot of time working on and talking about ‘self-driving IT’. That’s the intelligent, automated and dynamic provisioning of everything from compute and storage resources, to connectivity and bandwidth, to network services and applications.
It’s similar in many ways to the self-driving car phenomenon. Let’s take a look at some of the similarities.
In the long history of ‘autonomous vehicle technology,’ success wasn’t truly realized until the technology matured. Systems needed enough processing power to handle the massive volumes of data that had to be crunched in order to make real-time driving decisions – and they weren’t quite there.
But there was incremental progress along the way. Here are some of the more notable milestones.
- 1920’s – The “American Wonder” is briefly driven on the streets of New York City, guided by radio controls in a following car.
- 1961 – Developments by Stanford engineers led to creation of the Stanford Cart moon rover, the world’s first truly self-driving wheeled vehicle.
- 1987 –Mercedes-Benz develops a robotic van that can detect and avoid objects in the road.
- 2004-2013 – DARPA holds driverless car competitions in the Mojave Desert.
- 2014 – Major car companies roll out new autonomous driving features such as parallel parking, accident avoidance and lane change alerting.
Each of these small steps in innovation leveraged the technologies of the day, but processing power was the gating factor. The amount of processing power needed to mimic the functions of a human driver’s brain only became available recently. That high-end processing power opened the door to the advanced analytics and automated real-time response capabilities that driverless cars require.
Humans first truly vacated the driver’s seat in 2014. That’s when a Toyota Prius equipped with experimental driverless technology developed by Google was demonstrated in Nevada. The true, self-driving car had arrived.
Today, they can be found in Arizona, California, Texas, Michigan, Pennsylvania and Washington – but only in restricted test areas. Experts predict that they’ll be common on public roads in as little as a few years. At that point, we will have finally removed the complex yet mundane burden on humans to process and analyze all the data, and make all the right, split-second decisions that driving requires.
The parallel in the IT world is that lots of great ‘self-driving’ advances have been made over the years, particularly with the various forms of virtualization. Platforms such as SD-WAN, SDN and NFV are new sandboxes in which we can continue to evolve our automation capabilities.
Just like with cars, however, we’ll only be able to remove the burdens on IT teams – all the pressure-packed processing, analysis, and decisions – when the technology is powerful enough to handle the data.
Before the driverless car, we made incremental progress with things like anti-lock brakes, automated parallel parking, and lane change warnings. The same thing is happening in driverless IT. We’re offloading mundane tasks from IT and NetOps teams, and handling them with automation. Early examples include spinning up and down various application services, and automatically balancing traffic across networks. These changes free up people to focus on more strategic and creative initiatives – and drive further incremental progress.
The goal is a system that’s comprehensive enough to collect and analyze all of an organization’s network and machine data. That system also needs to have the smarts and processing power required to transform processed data into insights that can be instantly acted upon without human intervention.
This level of automation boosts organizations’ business agility and efficiency, creating real competitive advantages. It sounds an awful lot like our SevOne Data Platform.
So, I ask you, have we arrived on the doorstep of driverless IT? Computers can now replace humans in the driver’s seats of cars. Are we ready for them to do the same with seats in NOCs? I’d love to hear your thoughts.