Giorgio Grisetti is assistant professor at Sapienza University of Rome. He is member of the RoCoCo lab at La Sapienza since November 2010. He is also member of the Autonomous Intelligent Systems Lab. at Freiburg University headed by Wolfram Burgard where he worked as a Post Doc since 2006. His research interests lie in the areas of mobile robotics. His previous and current works aimes to provide effective solutions to mobile robot navigation in all its aspects: SLAM, localization and path planning. He was a PhD student at University of Rome "La Sapienza" in the Intelligent Systems Lab. His advisor was Daniele Nardi and he received his PhD degree in April 2006. His PhD thesis focused on SLAM using Rao-Blackwellized particle filters. In 2001, he received his M.Sc. degree in computer engineering, at the University of Rome.Current Teaching Activities
- Elective in Robotics, Module 3 Least Squares and SLAM , Spring/Summer 2013, La Sapienza
- Laboratorio di Programmazione , Spring/Summer 2013, La Sapienza
- Michael Ruhnke, 2007, Unsupervised Learning of 3D Object Models from Partial Views. Diplomarbeit, University of Freiburg. Published in proc. of ICRA 07. paper.
- Bastian Steder, 2007, Learning Maps in 3D using Attitude and Noisy Vision Sensors Diplomarbeit, University of Freiburg. Published in proc. of IROS 07. paper.
- Slawomir Grzonka, 2007, Look-ahead Proposals for Robust Grid-Based SLAM , University of Freiburg, Diplomarbeit. Published in Proc. of FSR-07. paper.
- Gian Diego Tipaldi, 2007, Speeding Up Rao-Blackwellized SLAM, ``La Sapienza'', Master's thesis. Published in Proc. of ICRA 06. Published in Journal of autonomous robots 06. paper;
- Andrea Censi, 2005, Scan Matching in the Hough Domain , ``La Sapienza'', Bachelor's thesis. Published in in Proc. of Int.~Conf.~on Robotics and Automation (ICRA), 2005. paper
- Luca Marchetti, 2005 A Comparative Analysis of Particle Filter based Localization Methods `La Sapienza'', Master's thesis. Published in Proc. of RoboCup Symposium, 2006.
- Sergio Lo Cascio, 2003, Design and Evaluation of Multi Agent Systems for Rescue Operations , `` La Sapienza'',Master's thesis. Published in Proc. of IROS 03.
During my activity, I was assistant of the courses of ''Fondamenti di Informatica 1'' (Basics of computer science) and Artificial Intelligence. For AI, I was also co-author with prof. Daniele Nardi of a set of notes on functional and logic languages. Additionally, I gave several seminars on state estimation, in the courses of AI and Vision and Perception. During and after the PhD, I advised several master thesis whose contents were often published in international conferences or workshops. This is a summary of those times:
- Elective in Robotics, Module 3 Least Squares and SLAM , Spring/Summer 2012, La Sapienza
- Doctoral Course on Least Squares and SLAM , Spring/Summer 2011, La Sapienza
- Fondamenti di Informatica 1 , Spring/Summer 2011, La Sapienza
- Robotics 2 , Winter 2010, Univ. of Freiburg
- Introduction to Mobile Robotics , Spring/Summer 2010, Univ. of Freiburg
- Robotics 2 , Winter 2009, Univ. of Freiburg
- Introduction to Mobile Robotics , Spring/Summer 2009, Univ. of Freiburg
- Introduction to Mobile Robotics , Spring/Summer 2008, Univ. of Freiburg
- Techiche di programmazione (in Italian), winter 2007, "La Sapienza", Rome
- Introduction to Mobile Robotics , Spring/Summer 2007, Univ. of Freiburg
- Introduction to Mobile Robotics , Spring/Summer 2008, Univ. of Freiburg
- Tutor for Intelligenza Artificiale , Summer 2006. ``La Sapienza'', 5 credits, approx 70 students.
- Tutor for Fondamenti di Informatica 1, Fall 2004 ``La Sapienza'', 5 credits, approx 70 students.
- Tutor for Intelligenza Artificiale, spring 2004 ``La Sapienza'', 10 credits, approx 70 students.
- Tutor for Fondamenti di Informatica Fall 2003 ``La Sapienza'', 10 credits, 100 students.
- Tutor for Fondamenti di Informatica , Fall 2002 ``La Sapienza'', 10 credits, 100 students.
- Tutor for Fondamenti di Informatica, Fall 2001 ``La Sapienza'', 10 credits, 100 students.
- Fondamenti di Informatica 1 , Spring/Summer 2012, La Sapienza
- Artificial Intelligence 2 , Spring/Summer 2012, La Sapienza
Research StatementIn the future, robots and autonomous devices will become more and morepopular and will finally be part of our everyday life. To achieve thisgoal, robot need to act intelligently and to offer useful services totheir users. Technically, the goal of an intelligent robot is to bringthe environment to a desired configuration by interacting with it. Toreach this goal the robot might perform complex sequences of actions.To calculate which actions to perform one needs to represent theenvironment at the level of abstraction appropriate for the specifictask. Our research focuses to develop techniques and models that canbe used in relevant classes of robotic applications, namely:autonomous navigation, object detection and object recognition. Themethods we developed have been used as building blocks of integratedrobotic systems which can explore the environment or that are used inindustrial applications. To speed-up the development of robotapplications we also contributed to middle-wares for robotics and weinvestigated how to compare different systems.Learning Models for Autonomous NavigationAcquiring maps with mobile robots has been deeply investigated in thelast two decades. This problem is known with the name of``Simultaneous Localization And Mapping'' (SLAM). It consists inestimating simultaneously the map and the position of a mobile robotwhile it moves and measures the environment with its on-board sensors.Within this context, we contributed with approaches that are able toconstruct metric models of the environment by using either filteringor smoothing approaches on graph representations.SLAM with Particle FiltersDuring the first period of our research we investigated particlefilters for SLAM. These filters represent the probability distributionover the possible states of the tracked system via a set of samples.In SLAM, every sample represents a potential robot trajectory. Weinvestigated how to accurately evolve this distribution of samples inorder to reduce their number, by placing them in the ``right''locations An opensource implementation of these algorithms (GMapping ) became well known to therobotic community.To improve the efficiency of these algorithms, we subsequentlyinvestigated on more compact representations that can exploit thesymmetries in the environment and in the SLAMprocess. These systems have beessuccessfully used in search and rescueapplications onhumanoids. They also have been extended tofixed-lag filters or to operate onmixed representations.Graph-based SLAMBased on the experience with particle filter SLAM, second phase of ourresearch we developed graph-based SLAM algorithms. As the namesuggests, these approaches model the problem as a graph. Each node ofthe graph represents a local map, and the position of this local mapwithin a global reference system. An edge between two nodes representsa spatial constraint between two local maps and is labeled with thespatial transformation obtained by aligning the two local maps. Thus,solving graph-based SLAM means:
- constructing the graph and the labels from the raw sensor data (graph construction) and
- find the locations of the local maps that better satisfies the constraints encoded in the edges (graph optimization).
To construct the graph we developed efficient algorithms to alignlocal maps made of 2D laser scans, on images andinertial data or on 3Dscans. To reject wrong alignments we developedalgorithms to efficiently compute the relative uncertainty between twonodes. To optimize the graph we developed an algorithm based onstochastic gradient descent.This algorithm relies on a tree parametrization of the graph to speedup the convergence. An open-source implementation has been madeavailable TORO ) and it isactually used by many research groups, for instance Willow Garage,Carnegie Mellon University, University of T'uebingen, University of Orebro,University of Parma and others.When the robot moves continuously in the environment the full SLAMproblem in its naive formulation becomes not feasible due to the largeamount of data that needs to be processed at each cycle. To this endwe investigated the possibility of dropping the less informativenodes. Furthermore, we approached theoptimization of these large scale problems by using hierarchicalrepresentations HOGMAN). Object Detection and Object ModelingTo perform complex tasks in unstructured environments, a robot shouldbe able to identify the known objects that it needs to manipulate. Inparticular, to manipulate an object a robot needs to find theinstances and the locations of this object in a scene starting from aknown model. In a recent work, we approached the problem of detectingknown objects in a 3D scene. The algorithmrelies on a compact representation of the three dimensional data(range images), and utilizes a variant of RANSAC to align a knownmodel on a portion of the scene. Subsequently, we developed analgorithm that learns models from the scenes in an unsupervisedway. This is done by aligning a portion of the scene onto another.The recurrent consistent portions are combined to form models. Thisproblem has substantial overlap with multi-robot SLAMAutonomous ExplorationMoving autonomously is an essential skill for any application thatinvolves mobile robots. To this end one needs a map of theenvironment. Thus, having a device which is able to autonomouslygather such an information is seen as highly relevant by the roboticcommunity. This task is called autonomous exploration.Traditional SLAM algorithms are substantially passive, in the sensethat they operate on a stream of data without controlling how thesedata are taken. However, it is known that the particular trajectoryfollowed by the robot has a substantial effect on the outcome of SLAM.We introduced an entropy based measure to drive the robot. This systemtakes into account the effects that a particular observation has onthe outcome of SLAM and chooses the one which maximizes theexpected information gain about the environment.Integration and Comparison of SubsystemsAn important aspect of our research is the systematic comparison andthe evaluation of different approaches in different scenarios.We proposed a metric to compare SLAMalgorithms which relates the quality of the map to the energy requiredto ``deform" it to obtain the ground truth. To quickly builddifferent applications involving mobile robots, we contributed to opensource systems with the realization of different components. Wefurthermore contributed to the design and implementation of anavigation system for small-size helicopters available at (http://www.openquadrotor.org).
- October 2010 Nomination for the best IROS paper award. M. Ruhnke, B.Steder, G. Grisetti, W. Burgard, Unsupervised Learning of Compact 3D Models Based on the Detection of Recurrent Structures.
- August 2010 Open Source achievement award from Willow Garage.
- April 2010 Best paper award at the International Conference and Exhibition on Unmanned Areal Vehicles. Samir Bouabdallah, Christian Bermes, Slawomir Grzonka, Christiane Gimkiewicz, Alain Brenzikofer, Robert Hahn, Dario Schafroth, Giorgio Grisetti, Wolfram Burgard, Roland Siegwart. Towards Palm-Size Autonomous Helicopters.
- April 2009 Best Paper award at ICRA 2009. Slawomir Grzonka, Giorgio Grisetti, Wolfram Burgard. Towards a Navigation System for Autonomous Indoor Flying.