By: Cindy Waxer
At Creating Revolutions, an employee affectionately nicknamed “Manuel Noriega” assembles the tiny components of a customer service paging device. Unlike other employees of the startup, Manuel works for hours, day in and day out, without bathroom breaks or healthcare benefits.
Meet today’s robot workforce. Manuel is a collaborative robot (or cobot) that’s helping Creating Revolutions build electronic tabletop devices for the restaurant industry. The startup didn’t always rely on a gunmetal grey robot arm to assemble its devices, which allow restaurant customers to text requests to busy wait staff. But faulty assembly was causing double-digit failure rates.
“The problem is you can’t efficiently repeat a specific process the exact same over and over again as a human being,” says Einar Rosenberg, CIO of Creating Revolutions.
With Manuel on the payroll, Creating Revolutions has reduced its product rejection rate to nearly zero. Changes to manufacturing processes can be made in real time for greater flexibility. And by cost-effectively increasing production rates, Creating Revolutions has managed to reduce its overhead by double digits. Employees initially bristled at the notion of sharing factory space with a cobot. But after assuring workers their jobs weren’t in jeopardy, Rosenberg says everyone now views Manuel as “part of the team.” In fact, the only thing separating Manuel from his human counterparts is a glass window pane.
Long gone are the days of bulky industrial machines in steel-metal cages. Nowadays, robots work side-by-side with human employees, often outpacing them in productivity for a fraction of their salary. Cloud giant Amazon relies on 45,000 robots to pick and pack orders across its fulfilment centers. Calls to Nanyang Technological University in Singapore are fielded by Nadine, a human-like robot programmed to display emotion. And the San Francisco robotics company Momentum Machines has developed a robot that cooks 400 burgers per hour.
In fact, according to Robotic Industries Association (RIA), in 2016, the North American robotics market broke all-time records for orders and shipments. During that year, 34,606 robots valued at approximately $1.9 billion were ordered in North America — a 10% uptick over 2015.
But as robots increasingly take on physical and cognitive tasks in the workplace, they promise to forever change the role of IT. Provisioning secure access, safeguarding data, programming complex robotic systems — they are all challenges facing IT leaders as robots usher in the next industrial revolution.
“The learning curve for introducing robots is pretty big,” warns Bob Doyle, director of communications at the Association for Advancing Automation. “There’s certainly training and education required to make sure that everything’s being done properly.”
It’s for this reason that Creating Revolutions turned to Hirebotics. Essentially a staffing agency for cobots, Hirebotics lets startups like Creating Revolutions rent cobots by the hour without any upfront capital expenditures. Hirebotics handles all programming, deployment and maintenance of the machine, while Creating Revolutions pays for only the hours it’s operable, nearly $5,000 per month.
“We’re a startup so we couldn’t afford a very expensive robot,” says Rosenberg. “And we didn’t have the expertise or the time to learn the software required to make the robot work for our needs. Economically, Hirebotics made sense.”
Although it took a few weeks for Hirebotics to program Manuel, Rosenberg says it would have taken Creating Revolutions’ IT team 10 times longer to withstand the “torture” of training the cobot to perform precision tasks. And because Manuel is connected to the cloud, Hirebotics can continuously monitor Manuel’s performance in real time for fast detection of glitches.
Reduced overhead, increased workloads
Complex programming isn’t the only obstacle IT leaders must overcome to reap the benefits of robotics. Consider, for example, Praxis Packaging Solutions. The contract packaging company owns 14 robots, including 13 Baxters — 300-pound, tomato-red, two-armed cobots; and one Sawyer, Baxter’s single-armed, nimbler sibling, both from Rethink Robotics.
With unemployment rates hovering around the 2% mark near Praxis’ western Michigan headquarters, Praxis CEO and president Richard King says, “We knew right away we’d be able to use Baxter to help fuel our growth.”
Today, Praxis’s band of Baxters perform a wide variety of repetitive tasks, from moving cardboard pieces to unpacking boxes. Programming Baxter is as simple and intuitive as “using an iPad,” according to Chris Hager, former director of IT at Praxis, who continues to advise the company as an IT consultant with Ferox Consulting. That’s because Baxter learns by demonstration — employees need only grab Baxter’s arms and move them, simulating the tasks it will be doing.
Despite its simplicity, enlisting Baxter for a series of tasks required some heavy lifting from IT. Praxis often runs multiple packaging orders in a single shift. As a result, the company needed to equip Baxter with a variety of attachments to allow for different uses. A 3D printer proved to be the answer. Today, Praxis can create a number of end effectors — devices that attach to the end of Baxter’s robotic arm to perform specific tasks.
But while a 3D printer has enabled Praxis to customize accessories for Baxter in real time and keep pace with new packaging requirements, King says, the IT team “still gets frustrated with” designing new 3D parts for its robots — a process that took months to master.
Making room for robotics
Praxis also had to make physical changes to its workspace to accommodate its cadre of cobots. Because its cobots are cage-free and seated next to co-workers, Praxis had to take some additional security precautions.
“We deployed some very small diagnostic cameras to certain areas [of the plant] that weren’t covered as a deterrent to prevent people from interfering with Baxter’s operation,” says Hager. These cameras also help the IT team diagnose performance issues. And a USB key with an embedded password is the only way workers can gain access to a machine.
Unlike Praxis, whose robots work side-by-side employees, Creating Revolutions’ Manuel “is kept in a separate and secure room,” says Rosenberg. “To access the robot, you need one of two keys, which only I and a manager have.” Only by physically swiping this key against a barcode located on Manuel can the machine be activated. Cameras also surround the area for constant and recorded monitoring.
But while cameras, passwords and barcodes protect robots from ill-intentioned employees, the question remains: What’s protecting workers from their 6-foot, heavyweight colleagues?
Before installing a robot in a common workspace, Doyle says, “A risk assessment needs to be done to determine exactly what a robot will be doing.” Within what proximity will a robot be working with humans? What safety features, such as sensors, will enable the robot to avoid physical contact with people? How can a robot be programmed to limit the force it uses to perform tasks?
By answering these questions, IT leaders can take steps to avoid personal injury. In fact, according to Universal Robots, whose machines are used by companies including Creating Revolutions, 80% of its cobots perform alongside human workers without safety caging, after an initial risk assessment is conducted.
Nevertheless, industry leaders are taking matters into their own hands. In February 2016, robotics experts published ISO/TS 15066, a set of safety standards for collaborative industrial robot systems in the workplace. Guidelines establish everything from minimum safety distances between man and machine to suggested maximum allowed speeds of operation. Interestingly, many of the standards are predicated on data regarding human pain thresholds.
“The fact is,” explained Carole Franklin, secretary of ISO/TC 299/WG 3, in a statement, “when robots work alongside humans, we have to be very careful that the application does not put a human in danger. Up until now, robot system suppliers and integrators only had general information about requirements for collaborative systems. ISO/TS 15066 is therefore a game changer for the industry. It gives specific, data-driven safety guidance needed to evaluate and control risks.”
The data deluge
Less easy to regulate is the sheer amount of data likely to be generated by cobots in the years to come. In the case of Creating Revolutions, Manuel provides a steady stream of performance and production data on components — bits and bytes that are automatically recorded every hour.
By gathering details on the number of units that Manuel produces in a given period of time, Rosenberg says, “We can combine all of that data and create realistic, very accurate forecasts of our production needs, such as how many humans are needed and what we need to focus on to produce at optimal levels.” From there, he says, Creating Revolutions can either “slow down certain processes” or “change what components the robot is making” to accommodate fluctuating demands.
As part of its machine-as-a-service model, Hirebotics also offers a software package that provides Creating Revolutions with reports, such as real-time charts for quick insight into exact production numbers.
Praxis, on the other hand, takes a more traditional approach to gathering the information generated by its cobots. “A lot of our data collection and analysis is done manually,” says Hager. That’s because most of the company’s robotic tasks run for just short periods of time, or are performed only once.
However, these are early days for cobots. At the moment, much of today’s proprietary hardware, such as sensors and conveyor belts, speak different languages, making it difficult for them to interact.
That’s changing as more and more companies discover how to connect cobots to other computer systems and applications. “At one time [security and privacy] weren’t really a concern,” recalls Dan Kara, a research director at ABI Research. “[Robots] were isolated devices that had little connectivity with anything else around them and didn’t collect a whole lot of information. That’s changed dramatically over time. Now what you have are robotics systems that capture huge amounts of data and interconnect with other devices in a working environment.”
All of which is giving rise to serious IT concerns. Says Kara: “Think of the amount of data that’s coming through a manufacturing environment or a warehousing environment – billions of pieces of data are created every day. For a CIO or a CTO, how do you manage all that data? How do you store that data? How do you make sure it’s secure? How do you derive value from that data?”
Preparing for smart factories
Helping to drive greater connectivity among cobots is innovative software such as Rethink Robotics’ Intera 5. The platform aims to help manufacturers integrate robots into factories in just a few hours. In addition to easier deployment, Intera 5-powered cobots can pick parts from a conveyor belt and then communicate with other computer-driven tools to guide the precise placement of these parts.
While it may not sound revolutionary, these interconnected devices are laying the groundwork for a smart factory — a manufacturing environment in which robots and computer-driven devices communicate and cooperate autonomously with one another via cloud computing and the internet of things.
Says Kara: “Now we’re starting to get these huge workflows within work cells where all of the devices are interconnected and talking to each other.”
Doyle points to cobot manufacturer FANUC and its Intelligent Edge Link and Drive (FIELD) system as another gateway to Industry 4.0 — tech parlance for a combination of digital transformations, including advanced robotics, artificial intelligence, sensors, IoT and data capture, that will revolutionize global manufacturing operations.
As more and more companies make use of tools such as FANUC, Kara says, “devices will not only generate data, but they’ll actually interact with other devices. It’s a really cool time; we’re moving towards the goal of Industry 4.0.”
For now, though, IT leaders are facing more immediate issues, namely security risks. Consider, for example, surgical robots. Surgical robots are increasingly being used to perform complex medical procedures, from removing gall bladders to stitching internal tissue. But in 2015, researchers at the University of Washington in Seattle managed to hijack a telesurgery robot, deleting and changing the order of commands it was receiving. The controlled experiment, conducted over a public network, highlights the increasing risks of hacking and other types of malicious attacks as bots achieve greater connectivity.
And then there’s the overall impact of robots on an IT environment. “When a company wants to bring in mobile robots, one of the biggest challenges IT leaders face is with IT infrastructure,” says Doyle. “That’s because the robot might be on a totally separate system.” For this reason, Doyle says, “it’s really important to ensure that the IT department and CIO are directly engaged in discussions from the beginning.” By predetermining a cobot’s connectivity to other devices, and the amount of data it will produce, companies can significantly reduce integration headaches.
Standardization is also a top priority for IT leaders with an interest in cobots. Take, for example, the OPC Foundation, whose mandate is to help enterprises maintain interoperability in their manufacturing and automation assets. Currently, the OPC Foundation is working on developing an industrial standard that will support interoperability among a wide range of manufacturing processes and equipment. By doing so, the organization hopes to ease data integration between machines and robots.
“It’s very important for organizations like the OPC Foundation to ensure that standards exist,” says Doyle. “If not, the biggest vendors are going to establish their own standards.”
From server rooms to factory floors
As industry consortiums and vendors hammer out new standards, IT leaders are rewriting their own playbooks. For instance, Hager recalls how his role at Praxis changed dramatically when the company purchased its first Baxter cobot. “[As IT leaders], we handle data collection and reporting and traditional IT jobs but we don’t necessarily get out there on the factory floor,” he says. “So [when we introduced Baxter], it was a transition from being in the backroom and managing IT operations to getting out on the floor and helping with production.”
To accommodate rising demands on IT, Hager says Praxis nearly doubled its IT team — a move that included hiring a senior network engineer and a helpdesk person. It’s a decision, he says, that allowed him to spend more time on the factory floor overseeing cobot operations.
In fact, the cobot revolution is even altering IT’s relationship with other departments. “In traditional automation environments, educated engineers only ask IT for help when things get tricky,” says Hager. “But with Baxter, training employees to work with cobots required more IT support.”
As global competition mounts, robots will continue to multiple across factory floors. In fact, the European Union is currently considering a proposal that would grant robots legal status as “electronic persons.” But before IT leaders can truly embrace cobots as co-workers, they’ll need to get a firm grip on security standards, data challenges and redefined roles.