Implementing robotics can be a slow, tedious process: Here are ways to make it easier

Think like a venture capitalist and make lots of small bets to see which one takes off, expert says. Pilot programs, the cloud and leases can also get a company started.

robotics.jpg

Image: Gorodenkoff/Shutterstock

Hands-free processing has become a mantra during the COVID-19 pandemic, with companies eager to execute processes that can avoid as much human contact as possible. This has created a market that is ripe for industrial robotics and process automation, which can reduce human handling, perform repetitive processes and can get products to market faster.

SEE: Hiring Kit: Video Game Designer (TechRepublic Premium)

Unfortunately, when bringing robots online, the process is anything but hands-free. Custom code must often be written to integrate robotics and industrial automation into existing applications. If the applications and systems have been customized over many decades, some of the code base that the automation must be integrated into might be "black box" code that no one understands and that might have been written more than 40 years ago. Then, there is the flip side of this: What if you determine that you want to incorporate robotics from a diverse group of vendors or replace robotics with a new solution that seems to fit the business better? Once again, you are likely faced with a large-scale manual integration to make it all work. 

The picture is clear. IT wants to implement robotics, the Internet of Things and automation, but the process can be long and fraught with error and risk. For these reasons, it's taking companies longer to integrate robotics and automation into their industrial operations. 

"Every time a company wants to interface new robotics, there's some level of integration effort that has to be done by software engineers," said A.K. Schultz, co-founder and CEO at SVT Robotics. "Once the company's business needs change and additional or different automation must be deployed, which is happening more and more frequently, most of that integration code has to be rewritten. That causes a scalability problem, and every single time you write the new code, it's all custom and must be maintained."

An example Schultz references is early-stage warehouse automation. "This automation often consisted of massive and expensive fixed systems that tended to focus on one big, trusted partner," he said. "That type of automation can cost hundreds of thousands to even millions of dollars, and the necessary process and enterprise software integration can take months or even longer, often due to an IT infrastructure bottleneck on the software side."

How can companies avoid situations like this? "First, I like to say that companies should think like a venture capitalist and make lots of small bets with the understanding that only a few will pay off," Schultz said. "Because of risk aversion, many companies will take a lot of time to do a thorough analysis and then place one big bet on a massive automation system. The problem with that is there's a good chance you didn't think of everything, so by placing a lot of small bets, you can figure out what technologies are actually good for your business."

SEE: The CIO's guide to quantum computing (free PDF) (TechRepublic)

Smaller companies have to be more cautious. One approach is to arrange "try and buy" pilots with vendors to see if proposed solutions address business needs and if they easily integrate with your enterprise code base. If a substantial amount of custom code must be written, or if the pilot experiences constant system breaks and repairs, it's probably wise to move on to something else.

Schultz also suggests that companies consider a cloud-based integration platform for robotics that supports standard interfaces and that can be more of a "plug and play" option. This enables companies to adopt different robotics and automation platforms, while also maintaining flexibility if they find a new solution that fits the business better. 

"The smaller, nimbler players that want access to robotics can lease them as needed, without having to spend $1,000,000 for a robot," Schultz said. "If a better technology comes along, and you want to be able to swap and upgrade for that, you can, because you don't have a capital investment that you're depreciating over multiple years."

Also see

By Mary Shacklett

Mary E. Shacklett is president of Transworld Data, a technology research and market development firm. Prior to founding the company, Mary was Senior Vice President of Marketing and Technology at TCCU, Inc., a financial services firm; Vice President o...