Localization with Dynamic Obstacles


The robots should operate in real environments where the uncertainty is a huge problem. Moreover dynamic obstacles, can give a false perspective of the environment to the robot that it may lead to unexpected results. We provided a way of getting the robot localized, using particle filter and a "move to localize" technique that decreases the uncertainty.


Alberto Quattrini Li, Marios Xanthidis, Jason O'Kane, and Ioannis Rekleitis.


[C66] Alberto Quattrini Li, Marios Xanthidis, Jason O'Kane, Ioannis Rekleitis. Active Localization with Dynamic Obstacles. In IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), 2016.

Experimental Comparison of open source Vision based State Estimation Algorithms

picture picture

In robot autonomy, estimating state and mapping, popularly known as Simultaneous Localization and Mapping (SLAM), is an active research field. There has been many works trying to solve SLAM problem, yet there exists no method that can be called robust such as it works in all environment and conditions. In this comparative study, we take recent and popular SLAM systems avaliable open source and try them on our datasets. These datasets are created as if they represent different environments with different conditions i.e. illumination, for multiple robots.

Creating a Mosaic of Coral Reefs using Drift-Nodes

picture picture picture picture

Coral Reefs are very important for the marine ecosystems and monitoring the coral reefs can be a very difficult task for the marine biologists. Providing a mosaic of the coral reef decreases significantly the labor for the experts while it provides a better and fast way to study better the coral reefs. We use simple and cheep to constract sensors, called Drift-Nodes, that been equipped with minimal cost of sensors (camera, GPS, IMU), they were deployed on the surface, above a coral reef in Barbados, and we used feature detection methods with feature matching to reconstract a mosaic.

Dynamically Efficient Kinematics for Hyper-Redundant Manipulators

picture picture picture picture

The Hyper-Redundant Manipulators is an old and classic problem in Control Theory and although it seems that it has been well studied in depth the past two decades only a few time efficient algorithms have been proposed. We propose an new algorithm for solving the Kinametics that potentially reduces the size of the problem and the time needed for the soluton exponentially by simplifying the kinematic structure. The idea is that we start from a simple kinematic structure by separating the manipulator into virtual sectors, where every sector uses a subset of its kinematic abilities and then in case of failure, the kinematic structure changes as described by the meta-controller, leading to a richer kinematic structure in the next step, and an increment of the dimensionality of the problem.


Marios Xanthidis, Kostantinos J. Kyriakopoulos, Ioannis Rekleitis


[C63] Marios Xanthidis, Kostantinos J. Kyriakopoulos, Ioannis Rekleitis. Dynamically Efficient Kinematics for Hyper-Redundant Manipulators. In The 24th Mediterranean Conference on Control and Automation, 2016.

Mapping Shipwrecks

picture picture

Historical shipwrecks are important for many rea- sons, including historical, touristic, and environmental. Cur- rently, limited efforts for constructing accurate models are performed by divers that need to take measurements manually using a grid and measuring tape, or using handheld sensors. A commercial product, Google Street View1, contains underwater panoramas from select location around the planet including a few shipwrecks, such as the SS Antilla in Aruba and the Yongala at the Great Barrier Reef. However, these panoramas contain no geometric information and thus there are no 3D representations available of these wrecks. This paper provides, first, an evaluation of visual features quality in datasets that span from indoor to underwater ones. Second, by testing some open-source vision-based state estimation packages on different shipwreck datasets, insights on open chal- lenges for shipwrecks mapping are shown. Some good practices for replicable results are also discussed.

## Participants A. Quattrini Li, M. Xanthidis, I. Rekleitis. ## Support We are gratefull for the generous support of a Google Faculty Research Award and of the National Science Foundation (NSF 1513203). ## References: [N9] A. Quattrini Li, A. Coskun, S. M. Doherty, S. Ghasemlou, A. S. Jagtap, M. Modasshir, S. Rahman, A. Singh, M. Xanthidis, J. M. O’Kane, I. Rekleitis. Vision-Based Shipwreck Mapping: on Evaluating Features Quality and Open Source State Estimation Packages. In MTS/IEEE OCEANS Monterey, 2016.
[N8] A. Quattrini Li, A. Coskun, S. M. Doherty, S. Ghasemlou, A. S. Jagtap, M. Modasshir, S. Rahman, A. Singh, M. Xanthidis, J. M. O’Kane, I. Rekleitis. On Understanding the Challenges in Vision-Based Shipwreck Mapping. In ICRA 2016 Workshop on Marine Robot Localization and Navigation, 2016.

Ongoing Projects

Fast Motion Planning for Redundant Systems

Thanks to the development of the robotics the last years more complicated robots have been presented executing complicated tasks, from mobile manipulators, cooperative systems and humanoids. But one of the biggest problems of using those systems effectively is the hard problem of the motion planning where even the best planners strugle to find solutions in reasonable time. Currently we are developing a new sampler for real-time planning that can be used by any motion planner and accelerates the solutions for those systems , dealing effectively with the curse of the dimensionality.