Program

08:35-08:40 Opening
08:40-09:20 Inivited Talk: Christopher Amato, (University of New Hampshire)
Multi-Robot Coordination Under Uncertainty With Limited Communication
(abstract) slides

The decreasing cost and increasing sophistication of robot hardware has created many new opportunities for teams of robots to be deployed to solve real-world problems. Although many algorithms have been proposed for multi-robot domains, the vast majority are specialized to match specific team or problem characteristics. Ideally, more general methods would exist for controlling multi-robot teams in a wide range of domains. In this talk, I will discuss such a general model for multi-robot coordination as well as methods for automatically generating solutions from a high-level problem description. These methods provide principled solutions that optimize control and communication decisions while considering uncertainty in outcomes, sensors and the communication channel. To demonstrate the scalability and effectiveness of these methods, I will show results from a cooperative beer delivery problem with heterogenous ground robots and a package delivery problem with teams of aerial robots.

09:20-10:00 Inivited Talk: Robert Fitch (University of Sydney)
Multi-Robot Active Perception for 3D Outdoor Object Recognition
(abstract) slides

Classifying objects in outdoor environments is a challenging problem in robotics and is fundamental to many applications, ranging from agricultural robotics to defence. Modern sensors and sophisticated perception algorithms can extract rich 3D textured information, but are limited to only the data that are collected from a given location or path. For example, object segmentation and classification is traditionally cast as a passive perception problem where data are generated as part of a disconnected navigation process and fed into a perception processing pipeline. In this talk, I will advocate instead for an active approach that closes the loop around perception and planning. I will discuss several recent results from my group in multi-robot coordination for viewpoint-dependent perception, probabilistic object representation using Gaussian process implicit surfaces, and non-myopic planning for active object classification.

10:00–10:30 Refreshment Break
10:30-11:15 Talks
10:30–10:45 William Bentz, Thomas Meitzler and Dimitra Panagou
A Redistribution Method for Multiple Energy-constrained robots in 2D Environments — key[pdf], [slides]
10:45–11:00 Aida Rahmattalabi, Jen Jen Chung and Kagan Tumer
D++: Structural Credit Assignment in Tightly Coupled Multiagent Domains [pdf], [slides]
11:00–11:15 Kyle Lund, Sam Dietrich and James Boerkoel
Robust Multi-Robot Scheduling  — key[pdf], [slides]
11:15-11:55 Inivited Talk: Gaurav S. Sukhatme (University of Southern California)
Life and Death: Online Decision Making for Information Gathering
(abstract)

Using examples from varied domains, this talk will argue that information gathering is a first-order activity for robots. Information gathering problems naturally lead to challenging online decision making problems. We will present recent results in two such problems in the single robot setting, and discuss the challenges associated with solving similar problems in the multi-robot case

12:00–14:00 Lunch Break
14:05-14:45 Inivited Talk: Gal Kaminka, (Bar Ilan University)
No robot is an island
(abstract)

We live in opportune times. The centuries-old dreams of creating intelligent, programmable automatons—robots—are becoming reality. Recent years are seeing dramatically growing interest in robotics, by scientists and practitioners alike. Robots—from the molecular scale to tank-size—seem to appear everywhere: in production lines, in the battlefield, in hospitals, in warehouses, in homes, in fields; on the ground, on water, and in the air.

In this talk, I argue that to accelerate and maximize the impact of robotics, robots should operate in teams, rather than in isolation. Moreover, I argue that effective robot teams must dynamically adjust their teamwork, instead of relying on rigid pre-planned coordination schemes. This is not a mere philosophical argument: I will present algorithms, data structures, and computational techniques for facilitating such teamwork, and discuss analytical guarantees and empirical results that demonstrate the effectiveness of these contributions in a variety of cooperative robot teams, from the molecular to the vehicle scale; robot teams which move in formations, explore urban areas, play soccer, and patrol Israel’s borders.

15:00-15:30 Talks
15:00–15:15 Qi Lu, Melanie Moses and Joshua Hecker.
A Scalable and Adaptable Multiple-Place Foraging Algorithm for Ant-Inspired Robot Swarms — key [pdf], [slides]
15:15–15:30 Sarah Tang and Vijay Kumar.
Translating Paths into Optimal Trajectories for Safe Coordination of Teams of Dynamic Robots — key [pdf], [slides]
15:30–16:00 Refreshment Break
16:00-16:40 Inivited Talk: Haluk Bayram (University of Minnesota)
Gathering Bearing Data for Target Localization
(abstract)

In this talk, I will consider the problem of controlling a mobile robot whose task is to cover an environment so as to localize one or more targets dispersed across the environment within a desired precision. We focus on a novel version of this general coverage problem in which the robot can collect only bearing measurements. Therefore, it must collect multiple measurements of a given set of locations to estimate the targets’ positions. The problem we study is to determine measurement locations and a coverage path for these locations. In doing so, the objective is to guarantee that the uncertainty in each position estimate does not exceed a given uncertainty level U while minimizing the data collection time. I will present an approximation algorithm and prove that its cost is at most 28.9 times the optimal cost while guaranteeing that the uncertainty is at most 5.5U. In addition to theoretical analysis, I will show the results in simulation and experiments performed with a directional antenna used for tracking invasive fish. I will conclude the talk with an overview of online data gathering research in our lab.

16:40-16:55 Talks
16:40–16:55 Jacques Saraydaryan, Fabrice Jumel and Olivier Simonin.
Modeling human flows from robots perception : application to navigation in dynamic environment — key [pdf], [slides]
16:55–17:10 Yinon Douchan and Gal Kaminka.
Using Reinforcement Learning to Improve Order Picking by Service Robots — key [pdf]
17:10-17:15 Closing
17:15 End