RobotRepublic— As part of a CEO interview series I’ve been working on for IIoT World’s, the industrial IoT zine I founded, I caught up with Sarcos CEO Ben Wolff. The topic: Industrial robotics. The outlook? Surprisingly .. human.
Check it out, below.
Lucian Fogoros: So Ben, it’s looking like the number of deployed industrial robots will hit 2.6 million units by 2019, at least according to the latest numbers.
That’s huge. It’s about a million units more the total industrial deployment in 2015, and that was a record breaking year. The majority of those deployments — about 70 percent — will take place in the automotive, electronics, metal and machinery sectors. But what about the rest? Who’s going to see growth first?.
Ben Wolff: Structured or predictable environments have always been more conducive to automation and robotic. Most of the robots deployed there are working in highly structured environments on tasks they can do better, faster or cheaper than humans.
Structured environments have always been more conducive to automation and robotic. Fewer variables and a limited problem set make the environment predictable, which is what you want (with the current technology) in these robots.
But that’s about to change. As advancements continue in sensors, machine learning and computing over the next few years, we’ll see a new category of industrial robots that are able to perform meaningful tasks in far less structured environments.
Construction, transportation, power, oil & gas, infrastructure maintenance and agriculture are going to be the big beneficiaries here.
In unstructured environments for industrial inspections, for example, there is a significant market for robots to be test equipment and infrastructure.
Because these are such time-consuming and, even, dangerous jobs for humans — and because they require such high rates of accuracy — they’re exactly the kind of jobs you want robots to take on.
Examples of such jobs include inspecting ballast tanks on ships, HVAC ducts in warehouses, wind turbines, nuclear power plant facilities and public infrastructure like sewers and subways …
Lucian Fogoros: But what about now, in 2017. How are robots reshaping industry this year?
Ben Wolff: We’re really living through a Golden Age of Robotics right now — especially as it relates computer vision, artificial intelligence, edge computing and cloud services, and the lowering co
sts of hardware and components. All those areas are fast maturing, which makes these technologies more cost-effective and viable for mass adoption than ever.
We’re reaching a tipping point, where soon robots will become a regular part of the workplace for a whole spectrum of industries.
In this year alone, we expect the numbe of robots to increase in industrial environments. In most cases, these robots will work alongside humans not to replace, but to augment human capabilities.
More are becoming connected to the cloud, too, a development that allows companies to store and analyze performance and environmental data over time. That’s the sort of data that companies need to begin to extrapolate trends and make predictions about the robots they’re utilizing.
It lets them look at every aspect of their deployment — from work environment to workload and how they’re doing in terms of helping them to create safer, more efficient workplaces.
Lucian Fogoros: Can you describe for me the different types of industrial robots we’re currently seeing in these environments?
Ben Wolff: Well , the majority of robots deployed in industry right now are fixed in place. And as I said, they’re really just performing just a couple of limited number of highly repetitive functions.
Robots on an automotive assembly line are a great example of this.
These are robots that are capable of performing a task faster, more precisely and a lot cheaper than a human can, but only within these highly structured and predictable environments.
That’s because these machines are still somewhat limited in their ability to recognize anomalies or variations and correct their course the way humans did, and that’s both a hardware and software limitation, right now.
Lucian Fogoros: According to an IOActive report that was published in February 2017. full ecosystem of robots presents a huge attack surface with numerous options for cyberattacks. Knowing that, what percentage of your resources do you allocate for securing your robots?
Ben Wolff: From the initial conceptual design through commercialization of a product, we’re incorporating the same security architecture into our commercial and industrial products that we’d do for robots in critical sectors, like public safety and defense.
Lucian Fogoros: What are the most critical security issues that can arise once you’ve made a data process set Internet-accessible?
The most critical security issues would include a hacker taking control of a robot. It’s the same sort of scenario researchers have demonstrated and warned about with connected cars and hackers.
Robots and connected cars have a lot in common. The way you mitigate this problem is by having discrete systems, some of which are impossible to access or remotely access over a network.
Lucian Fogoros: What about industrial robotic solutions? What’s on the horizon.
Ben Wolff: It’s a good bet that greater technical capabilities will result in greater utility. Robots will increasingly move from being single purpose to multi-purpose, capable of performing a variety of different tasks. They’ll be far more agile and adaptable, too. And because they can do more, you can generate a return on investment more quickly.
Also, robots are going to become increasingly collaborative. That means they’ll be able to safely work side-by-side with humans, and without having to physically separate robots from the human workspace.
Finally, we’ll see more and more robots working in partnership with a human operator. That way, the human’s wisdom, judgment and experience is always in play, but it’s paired and augmented by a machine’s precision, strength and robustness. What you end up is something truly useful, a man-machine pair to perform tasks that neither man nor machine could perform independently.
For RobotRepublic, I’m Lucian Fogoros.
A version of this story ran on IOTT World. Read it here.