Robots are perfectly suited for certain jobs, especially when they are carefully engineered to perform them. Without proper synergy, however, using robots can be problematic—and sometimes disastrous.
On the positive side, in five separate Amazon warehouses located around the world, robots pack up to 700 products an hour. That is four to five times faster than the humans who would otherwise be performing the repetitive, tiring task.
According to Chris Quick, founder and CEO of RealBotics, Inc., robots are usually a good fit in tasks that require precision and repetition. Determining the appropriate application for a robot begins with a series of questions regarding the process at hand.
For example, consider everything required for a robot to wash windows on the side of a building.
“At first you start thinking about a robot arm spraying water on a window and wiping it off,” Quick says. “Maybe you also think about the programming required to teach the robot [to wash] each window or the sensor that might detect a glass surface from a non-glass surface.”
However, he says there are several questions that need to be considered before implementing such a solution, and not all of them are obvious. Quick listed examples: “How will the robot clean off its squeegee? Where does the water come from? How is the robot powered? How will the system deal with wind, heat and rain? What happens if a window is open? How does the robot move from one window to the next, one side of the building to the next? From one building to the next? How much time does it take the system to clean a window? How does it behave when close to the top floor or bottom floor of the building? What happens if the system loses power or encounters an error condition?”
Robots make sense for completing high-volume and high-precision tasks, as well as those in dangerous environments. In such cases, the costs for engineering and integration are easier to justify.
However, robots are not right for every situation. At times, the cost is prohibitive. Furthermore, they simply cannot perform certain tasks.
“Many [robotics] projects are never completed due to unforeseen complexities or substantially underestimated timelines,” Quick says. “Others become operational but do not perform as envisioned and are quickly abandoned.”
In a less technical example, Honda developed a humanoid robot to be a guide at Tokyo’s Miraikan museum. Problems developed when the robot could not distinguish between when people were making random arm motions or raising a hand to ask a question. In this case, the use of humanoid robots was premature, as designers did not incorporate people’s unexpected interactions.
The tasks may be suitable for a robot to perform, but the robot must be properly designed for accuracy and functionality, and the designers must account for all possible scenarios. This includes proper coding, development, motion and inclusion of any components and materials the robot needs to complete the desired tasks.
After this is achieved and the robot is properly designed, a cost-benefit analysis must be performed to justify its use. This approach helps to guarantee a positive outcome.
“Identifying a company’s need (and appetite) for robotic automation starts with developing an understanding of the problem and ensuring the primary drivers can be quantified. If a justifiable solution can be well-defined, the project is much easier to sell up through an organization,” says Thomas D. Pyper Jr., ASG sales and business development manager at Applied Manufacturing Technologies. “Prior experience with automation and the staff to support it are also important factors. If the problem can be solved with a robotic solution and the customer is willing to commit the time and resources to move the project forward, you have the makings of a successful project.”