Contents
25 November 2020
Vol 5, Issue 48
Vol 5, Issue 48
Editorial
- Robots are not immune to bias and injustice
Social constructs drive human-robot interactions; robotics is thus intertwined with issues surrounding inequity and racial injustices.
Focus
- Marine animal tracking with classical and emerging localization algorithms
Localization algorithms applied to acoustic tags for tracking marine animals can also be used to localize marine robots.
- Deep diving with Clio
An autonomous underwater vehicle, named Clio, can sample ocean basin–scale biogeochemistry at depths up to 6000 m.
- Spray-on magnetic skin for robotic actuation
A minimalist robot construction strategy offers versatility and compatibility in actuating diverse objects on demand.
Research Articles
- Mobile robotic platforms for the acoustic tracking of deep-sea demersal fishery resources
Mobile robots enhance traditional acoustic tracking methods to study the spatiotemporal behavior of deep-sea fishery resources.
- Revealing ocean-scale biochemical structure with a deep-diving vertical profiling autonomous vehicle
Autonomous sample return and high-resolution depth control enable global-scale biochemical mapping of marine genomics and proteomics.
- Deep learning can accelerate grasp-optimized motion planning
Warm-starting grasp-optimized motion planning with deep learning reduces computation time by two orders of magnitude.
- An agglutinate magnetic spray transforms inanimate objects into millirobots for biomedical applications
Millirobots with magnetically drivable surfaces display a range of locomotive abilities and programmable behavior.
About The Cover
ONLINE COVER Smooth Manipulator. Robot picking for commercial applications requires smooth robot arm motions from a variety of configurations to be computed quickly and efficiently. Ichnowski et al. have developed an algorithm for robot picking that considers jerk limits to produce smoother trajectories and uses deep learning to speed up the computation of those trajectories. The algorithm, which was validated on a UR5 robot, reduced the computation time of a preexisting motion planner from 25 seconds to 80 milliseconds. This month's cover is a photograph of a UR5 robot picking objects. [IMAGE CREDIT: ADAM LAU/BERKELEY ENGINEERING]
