Listen to this article

The Massachusetts Institute of Technology (MIT) recently released a study called “The State of Industrial Robotics: Emerging Technologies, Challenges, and Key Research Directions” that described challenges of adopting and using industrial robotics. It also made recommendations about how to move the industry forward.

Mainly focused on industrial robotics in Europe, the report detailed findings from interviews with robot manufacturers, original equipment manufacturers that use robotics technologies, and industrial research institutions. The report used the International Federation of Robotics’ definition of industrial robots, saying they are “automatically controlled, reprogrammable multipurpose manipulators programmable in three or more axes.”

You can read the full report here (PDF). The report was produced as part of the MIT Work of the Future initiative. Worker resistance has not seemed to play a big role in slowing adoption of industrial robots, according to MIT. This is partly because of worker protections in Europe that limit layoffs and partly because leading companies have recognized that involving workers in decisions on how to redesign manufacturing lines results in both better technical decisions and greater acceptance of new technology.

Challenges holding back industrial robots

High cost of integration: While robot hardware has become cheaper, the cost of the robot is a small fraction of the overall price tag of introducing automation to a manufacturing line. The study cited one large, well-known robotics integrator that said, “robots are now inexpensive, but integration is not.” And not only can the initial integration process be long, arduous, and expensive, even small changes to a manufacturing line often require bringing in an integrator to redesign and repurpose robot workcells to meet the specifications of new tasks.

According to the report, many companies cited the high costs of integration as a barrier to adoption. Large manufacturers using industrial robots can get weighed down attempting to integrate existing 15- to 20-year-old technologies and infrastructures with new robotics technologies on the same manufacturing line. On the other hand, SMEs with production processes that run at a smaller scale, often find integration costs to be prohibitive or unjustifiable due to their smaller production lot sizes.

Lack of standards: The absence of standardization in many aspects of robotics made it difficult for some companies to exploit the full potential of promising technologies or to incorporate new types of robots into manufacturing lines. Some companies developed home-grown systems and capabilities over the years that they would prefer to continue to use. Others, often smaller companies, are designing manufacturing lines or purchasing new technologies from the ground up. Each type of robot might require knowledge of different programming languages, interfaces, or communication protocols for use. Standardization should cover all of these considerations, besides safety standards and hardware.

“Standardization is not as easy as one thinks,” according to a large robotics company cited by the study. “There are many experts to listen to. It must go hand-in-hand with which technologies are capable of what things and an understanding of the processes they’re part of.”

Inflexibility: The fact that current robotics technologies cannot always be quickly repurposed limited the potential use cases of robots as well as experienced line workers’ ability to leverage their deep domain knowledge to improve manufacturing processes through direct repurposing of the robots.

Flexibility was also considered important for increasing the applicability of robotics to different levels of production, including high-mix and low-volume production; faster integration and reintegration times; re-configurable workcells and manufacturing lines; reducing the factory footprint by allowing manufacturing of multiple products along a single line; and enabling the reusability of robots.

Better balance of speed and safety: Collaborative robots are inherently safer, but uptake remains limited in some environments. The report said large companies favor speed over integration with humans, especially since cobots are not optimal for certain tasks. A balance among safety, speed, and versatility would encourage investment in cobots.

Data protocols: Better data infrastructures could help companies make “smart data from big data,” as a large multinational put it. As it stands, companies are resistant to investing in these infrastructures, as a clear benefit does not currently exist.

Improvements to enabling technologies: Extreme robustness, faster integration, and removal of technological bottlenecks related to sensing, perception, and gripping are critical to encouraging investment in and the mass adoption of automation. For example, the study said that while vision technologies are promising for ensuring the safety of robots and reducing quality check costs, they are still a long way from being optimal.

“Our research shows that vision systems and corresponding algorithms often performed well in research settings or other well-controlled settings, but they broke down, in different ways, in actual factories. Without rock-solid reliability, recent advancements in algorithms (such as deep learning-based architectures) are not robust enough to be used.”