In my previous two posts in this series, I've essentially argued both sides of the same issue. In the first, I explained why deep learning is not a panacea, when machine learning systems (now and likely always) will fail, and why deep learning in its current state is not immune to these failures. In the second post, I explained why deep learning, from the perspective of machine learning scientists and engineers, is an important advance: Rather than a learning algorithm, deep learning gives us a flexible, extensible framework for specifying machine learning algorithms. Many of the algorithms so far expressed in that framework give orders of magnitude-level improvement on the performance of previous solutions. In addition, it's a tool that allows us to tackle some problems heretofore unsolvable directly by machine learning methods.

Researchers have unveiled an octocopter that attaches asphalt shingles to roofs using a nail gun and'virtual switch'. This robotic carpenter's software measures the force needed to compress the point of the nail gun and a human operator activates a virtual switch when the robotic carpenter is in place. Researchers have unveiled its roofing octocopter that attaches asphalt shingles to roofs using a nail gun and'virtual switch' The roofing octocopter is the brain child of researchers at the University of Michigan, which is part of the movement to shift humans from'dull, dirty and dangerous job' into'cleaner, safer and more interesting jobs'. 'For me, the biggest excitement of this work is in recognizing that autonomous, useful, physical interaction and construction tasks are possible with drones,' said Ella Atkins, a professor of aerospace engineering and robotics. The team used a system of markers and stationary cameras to enable the drone to precisely locate itself in space and the system was also used to tell it where to place the nails.

The FAA has warned against equipping your drone with weapons such as flamethrowers and handguns. But can a nail gun really be considered a weapon -- that is, outside of Quake? Let's hope not, because roboticists at the University of Michigan have made a roofing drone that uses that tool to autonomously nail shingles into place. In a video shot in UM's special drone testing habitat, the craft flies up, approaches its bit of roof, and gingerly applies the nail gun before backing off and doing it a couple more times. It's very much just a tech demonstration right now, with lots of room to improve.