Inside Unmanned Systems

DEC 2016-JAN 2017

Inside Unmanned Systems provides actionable business intelligence to decision-makers and influencers operating within the global UAS community. Features include analysis of key technologies, policy/regulatory developments and new product design.

Issue link: http://insideunmanned.epubxp.com/i/763107

Contents of this Issue

Navigation

Page 50 of 59

51 unmanned systems inside December 2016/January 2017 ENGINEERING. PRACTICE. POLICY. Software and data from sensors such as ac- celerometers, barometers, gyroscopes and GPS also help the drone compensate for how wind and spraying can push it around. Dahlstrom said his previous experience in the software industry is key to his drone work now. "I've often ref lected on why nobody else tried doing something similar to this beforehand," Dahlstrom said. He recalled that at an event with other drone companies, others would say "that drones can't do that. Because my back- ground is in software, I see drones as f lying computers, and see them as able to do things that maybe other people don't see." "For instance, some hobbyist quadcopters have a button where you can push a button and it will execute a f lip," he said. "To do that manually takes hundreds of hours of f light time to do skillfully. Once you put it into code, it's a no-brainer—it can do it consistently and repeatably and accurately." "We're doing things drones were never in- tended to do," Dahlstrom said. "They were intended to f ly far and fast and not have any- thing attached to them. Ours are f lying close to a surface, and slow, and tethered." Autonomous Future The Apellix system is only semi-autonomous—it can't do 100 percent of the painting. Instead, pi- lots remote control the drone, which is throttled to a maximum speed of 3 mph. "You press a but- ton and the software moves it to the optimal dis- tance from the surface, and you move the joystick to paint," Dahlstrom said. Still, "this computer-assisted flight can achieve about 80 percent of the benefits of a fully autono- mous system and get to the market much quicker and with less expense," Dahlstrom said. "Even if it was fully autonomous, it'd still need a person there to monitor the system and take over should something go wrong. But we do aim for it to be fully autonomous in the future." Although Apellix's cautious pace of develop- ment "might be slower than I think anybody would like, I think getting it right is really im- portant," Salvatore said. "Getting a consistent spray is important, and a stable drone is really, really important." A large part of painting is not simply the act of applying paint, but also preparing surfaces beforehand. "The prep work still needs to be done by people," Dahlstrom said. "As the technology improves, maybe some of that prep work can be done by drones as well, such as washing or sandblasting. Right now if the drone is paint- ing a surface and it comes to an area that needs prep, it will skip it in continue spraying. Once the job is mostly complete, then you can send people in, do prep work, and send the drone back to touch up those areas." Apellix has nondisclosure agreements signed with some of the world's largest paint and spray equipment manufacturers, Dahl- strom said. Apellix hopes to have a commer- cially available spray-painting drone the first half of 2017. "We get contacted quite often with poten- tial ideas for the drone, anything from paint- ing power transmission towers to cell phone towers and even wind turbines," Dahlstrom Researchers in Singapore say their ground-based PictoBot robot will be able to paint high interior walls faster than two painters combined. "(PICTOBOT) IS A PLATFORM technology that you can mount something else on for other purposes—for instance, you can paint a very thin layer of cement to make a surface very smooth." I-Ming Chen, director of Nanyang Technological University's Robotics Research Center, Singapore

Articles in this issue

Archives of this issue

view archives of Inside Unmanned Systems - DEC 2016-JAN 2017