Flocking control is a challenging problem, where multiple agents, such as drones or vehicles, need to reach a target position while maintaining the flock and avoiding collisions with obstacles and collisions among agents in the environment. Multi-agent reinforcement learning has achieved promising performance in flocking control. However, methods based on traditional reinforcement learning require a considerable number of interactions between agents and the environment. This paper proposes a sub-optimal policy aided multi-agent reinforcement learning algorithm (SPA-MARL) to boost sample efficiency. SPA-MARL directly leverages a prior policy that can be manually designed or solved with a non-learning method to aid agents in learning, where the performance of the policy can be sub-optimal. SPA-MARL recognizes the difference in performance between the sub-optimal policy and itself, and then imitates the sub-optimal policy if the sub-optimal policy is better. We leverage SPA-MARL to solve the flocking control problem. A traditional control method based on artificial potential fields is used to generate a sub-optimal policy. Experiments demonstrate that SPA-MARL can speed up the training process and outperform both the MARL baseline and the used sub-optimal policy.

This research focuses on soft robotic bladders that are used to monitor and control the interaction between a user's head and the shell of a Smart Helmet. Compression of these bladders determines impact dissipation; hence the focus of this paper is sensing and estimation of bladder compression. An IR rangefinder-based solution is evaluated using regression techniques as well as a Neural Network to estimate bladder compression. A Hall-Effect (HE) magnetic sensing system is also examined where HE sensors embedded in the base of the bladder sense the position of a magnet in the top of the bladder. The paper presents the HE sensor array, signal processing of HE voltage data, and then a Neural Network (NN) for predicting bladder compression. Efficacy of different training data sets on NN performance is studied. Different NN configurations are examined to determine a configuration that provides accurate estimates with as few nodes as possible. Different bladder compression profiles are evaluated to characterize IR range finding and HE based techniques in application scenarios.

Uncle Sam doesn't really want a gangly 18-year-old soldier to stand guard outside the gate of a military base, rather he wants a wide-area motion imagery (WAMI) system that provides surveillance, reconnaissance, and intelligence-gathering using specialized software and camera systems to detect and track hundreds of people and vehicles all at the same time over a city-sized area. Uncle Sam doesn't really want a blurry eyed, half asleep and distracted human pilot flying in circles trying to find camouflaged bad guys on the ground, rather he wants a multispectral system, that can see things invisible to human eyes, consisting of four high-definition cameras covering five spectral bands; a three-color diode pump laser designator and rangefinder; laser spot search and track capability; automated sensor and laser bore sight alignment; three-mode target tracker., and MTS sensors that offers multiple fields of view, electronic zoom, and multimode video tracking. Uncle Sam doesn't really want more spies in trench coats that lurk in dark corners vaping, rather he wants persistent surveillance systems that collect and integrate data from specific geographic areas with data on activities that happened there at specific dates and times. This capability requires a spatiotemporal analytic method to recognize trends and patterns from large, diverse data sets. These data sets identify activities: events and transactions conducted by entities (people or vehicles) in an area, while documenting patterns of life and alerting to unusual events.

This is a guest post. The views expressed here are solely those of the author and do not represent positions of IEEE Spectrum or the IEEE. Olympic gymnast Simone Biles has a signature move that is named after her because she is the only woman on earth capable of performing it. The move starts as a layout double flip, but more than halfway through suddenly develops a twist that rotates her body through an extra 180 degrees to land face first. The only visible source of this sudden change in rotation is a small motion of one hand as her arm goes from straight to bent.

MIT researchers have built a wearable sensor system that automatically creates a digital map of the environment through which the wearer is moving. The prototype system, described in a paper slated for the Intelligent Robots and Systems conference in Portugal next month, is envisioned as a tool to help emergency responders coordinate disaster response. In experiments conducted on the MIT campus, a graduate student wearing the sensor system wandered the halls, and the sensors wirelessly relayed data to a laptop in a distant conference room. Observers in the conference room were able to track the student's progress on a map that sprang into being as he moved. Connected to the array of sensors is a handheld pushbutton device that the wearer can use to annotate the map.

Robotic killers that detect feral cats, spray their fur with poison and rely on them to essentially lick themselves to death have been deployed in the Australian desert for the first time. Feral cats are one of the biggest threats to many of Australia's endangered species, killing millions of animals every day throughout the country – and controlling them has proved difficult. It took John Read, an ecologist seven years to invent and produce four of the "grooming traps". After extensive testing, he has switched on the first one in a nature reserve in south-west Queensland. "Cats are hard-wired to hunt," Read said.

A team of researchers at the MIT Computer Science and Artificial Intelligence Laboratory (CSAIL) led by Li-Shiuan Peh has come up with a new infrared depth-sensing system. The new system, which works outdoors as well as in, was built by attaching a US 10 laser to a smartphone, with MIT saying the inexpensive approach could be used to convert conventional personal vehicles, such as wheelchairs and golf carts, into autonomous ones. Inexpensive rangefinding devices, such as the Microsoft Kinect, have been a great help to robotics engineers. Using the off-the-shelf product that relies on an infrared laser to measure distance, they allow for rapid prototyping and the ability to create robots that can sense and navigate in their environments without having to constantly reinvent the necessary technology. Unfortunately, Kinect and similar infrared-based systems tend to be a bit fussy when it comes to ambient light conditions.

This paper applies the Mixture of Gaussians probabilistic model, combined with Expectation Maximization optimization to the task of summarizing three dimensional range data for a mobile robot. This provides a flexible way of dealing with uncertainties in sensor information, and allows the introduction of prior knowledge into low-level perception modules. Problems with the basic approach were solved in several ways: the mixture of Gaussians was reparameterized to reflect the types of objects expected in the scene, and priors on model parameters were included in the optimization process. Both approaches force the optimization to find'interesting' objects, given the sensor and object characteristics. A higher level classifier was used to interpret the results provided by the model, and to reject spurious solutions.

This paper applies the Mixture of Gaussians probabilistic model, combined with Expectation Maximization optimization to the task of summarizing three dimensional range data for a mobile robot. This provides a flexible way of dealing with uncertainties in sensor information, and allows the introduction of prior knowledge into low-level perception modules. Problems with the basic approach were solved in several ways: the mixture of Gaussians was reparameterized to reflect the types of objects expected in the scene, and priors on model parameters were included in the optimization process. Both approaches force the optimization to find'interesting' objects, given the sensor and object characteristics. A higher level classifier was used to interpret the results provided by the model, and to reject spurious solutions.

This paper applies the Mixture of Gaussians probabilistic model, combined withExpectation Maximization optimization to the task of summarizing threedimensional range data for a mobile robot. This provides a flexible way of dealing with uncertainties in sensor information, and allows theintroduction of prior knowledge into low-level perception modules. Problemswith the basic approach were solved in several ways: the mixture of Gaussians was reparameterized to reflect the types of objects expected in the scene, and priors on model parameters were included in the optimization process. Both approaches force the optimization to find'interesting' objects, given the sensor and object characteristics. A higher level classifier was used to interpret the results provided by the model, and to reject spurious solutions.