The Robot Art Competition is a fascinating blend of art and technology, challenging engineers to create robotic systems that can produce artworks using physical brushes and paint. The winners of the third annual competition highlight the growing sophistication and variety of machine-generated artwork.

The competition, founded by Stanford educated mechanical engineer Andrew Conru, is primarily interested in how well robotic engineers can develop new mechanical painting devices. In fact, the only real guiding limitation in the rules is that, "Paint/color must be applied with one or more physical brushes by a robotic system."

Within this constraint, the 19 teams that entered the 2018 competition all devised novel ways for their robots to create final art works. Some teams created robot arms that can mimic the movements of human artists, while others developed more complicated input processes, with their robot artists directed by various software algorithms. Some teams even went so far as to use AI systems to get their robots to generate a completely original image without a source photo or image for inspiration.

The winners were selected by a mix of public voting and professional judging. The professional art critics judging the winning entries were directed to evaluate the artworks based on various criteria: overall originality and aesthetics, "painterly" ability, and technical contribution. The top 10 entries took home a share of US$100,000.

The first-placed robot this year, taking home the $40,000 lion's share of the prize pool, is called CloudPainter. Developed by independent American roboticist Pindar Van Arman, CloudPainter came third in last year's competition and is rapidly evolving in sophistication from year to year. This is one of the more autonomous generative systems in the competition, and Van Arman notes that more teams have started incorporating AI into their robotic systems across each subsequent year of the competition.

"While AI was unusual when the contest began, it has since become one of the most important tools for the robots," writes Van Arman. "Many of the top entries, including mine, Hod Lipson's, and A Roboto used deep learning to create increasingly autonomous generative art systems. For some of the work it became unclear whether the system was simply being generative, or whether the robots were in fact achieving creativity."

Second place and $25,000 went to a series of works from Columbia University's Creative Machines Lab. The robot, dubbed PIX18, won the competition last year, and its efforts this year demonstrated a growing sophistication in re-interpreting existing art using impressively nuanced brushstrokes.

Take a look through our gallery for a look at all the winners in this year's Robot Art Competition.

Source: Robot Art

View gallery - 33 images