Scientists aren’t guaranteeing perfection, but they do think that greater efficiency is a reliable product of practice,–20 percent greater efficiency, to be precise. Researchers at the University of Colorado in Boulder report that when they studied the ways that test subjects learned particular arm-reaching movements using a robotic arm, they found that even after a task had been learned and the corresponding decrease in muscle activity had reached a stable state, the overall energy costs continued to decrease, which seems strange. A University of Colorado news release reports that at the end of the task, (which included more than 600 repetitions of reaching movements) the net metabolic cost as measured by oxygen consumption and carbon dioxide exhalation had decreased by about 20 percent. The experiment studied people who used a handle on a robotic arm, similar to a joystick, to control a cursor on a computer screen, moving it from a set position to reach for a target on the screen. The experiment was set up so that the test subjects had to exert more energy in some reaching movements when the robotic arm created a force field, making subjects “push back†as they steered the cursor toward the target. Measurements of oxygen consumption and surface electromyographic data from six upper limb muscles suggested that the subjects learned the task by not only cutting down on errors, but also by reducing effort. How did they do that? The answer, the researchers say, must be that efficient movements involve both efficient biomechanics and efficient neural processing, otherwise known as thinking. Read more from the University of Colorado.