Student Projects

DISAL offers a range of student projects its three main areas of expertise: Distributed Robotic Systems, Sensor and Actuator Networks, and Intelligent Vehicles. For more information about supervision guidelines and different types of student projects please refer to this page.

Fall Semester 2022-2023

Automatic Design of Behaviors for Khepera IV Robots Emulating Automotive Vehicles

 

With the increased deployment of mobile robotics in a multitude of real-life scenarios, robots must face more and more complex situations. However, every robotic platform needs to be carefully programmed to fulfill its task, which, today, is often still done manually. Such programming operation usually represents an intense time investment and requests high expertise to eventually achieve the targeted performances in the real deployment, even more so if the application involves a multi-robot system. There are multiple approaches for the automatic generation of controllers (or specific components), whereas the nature of the control components ranges from Probabilistic Finite State Machines (PFSM) [1] to Artificial Neural Networks (ANNs) [2].  However, to automatically generate any of these controllers, powerful machine-learning techniques are necessary.

This project aims to automatically generate basic behaviors which can then be leveraged using behavioral arbitrators for more complex scenarios. Starting with a thorough literature research on the state-of-the-art of automatically generated basic behaviors such as [3], the most promising option will then be implemented using Webots, a high-fidelity, open-source robotics simulator. In particular, a library of basic behaviors will be generated for Khepera IV robots. The combined effectiveness of the resulting behavior-based control architecture will be demonstrated in a realistically simulated traffic scenario, with the Khepera IV robots emulating automotive vehicles. The expected effort and outcome of this work will be adjusted based on the type of the project.

Recommended type of project: Semester project / Master project

Work breakdown: 30% theory, 40% software, 30% experimentation       

Prerequisites: broad interest in machine learning, control algorithms and robotics, good command of C/C++

Keywords: machine learning, robot control, automatic design and optimization

Contact: Cyrill Baumann

References:

[1] C. Baumann and A. Martinoli, “A Noise-Resistant Mixed-Discrete Particle Swarm Optimization Algorithm for the Automatic Design of Robotic Controllers” to appear in IEEE Congress on Evolutionary Computation (2022)
[2] Duarte M., Sancho Moura O., and Anders Lyhne C., “Evolution of Hybrid Robotic Controllers for Complex Tasks” Journal of Intelligent & Robotic Systems (2015) 78: 463
[3] Jones S., Studley M., Hauert S., Winfield A., “Evolving behaviour trees for swarm robotics” International Symposium on Distributed Autonomous Robotic Systems 2016

Automatic Design of Flexible Behavioral Arbitrators for Khepera IV Robots

With the increased deployment of mobile robotics in a multitude of real-life scenarios, robots must face more and more complex situations. However, every robotic platform needs to be carefully programmed to fulfill its task, which, today, is often still done manually. Such programming operation usually represents an intense time investment and requests high expertise to eventually achieve the targeted performances in the real deployment, even more so if the application involves a multi-robot system. There are multiple approaches for the automatic generation of controllers (or specific components such as behavioral arbitrators), whereas the nature of the control components ranges from Probabilistic Finite State Machines (PFSM) [1] to Artificial Neural Networks (ANNs) [2].  However, to automatically generate any of these controllers, powerful machine-learning techniques are necessary.

This project aims at exploring behavior trees [3], structure that can implement both serial and parallel behavioral architectures and therefore lie between PFSM- and ANN-based behavioral arbitrators. Leveraging a Mixed-Discrete Particle Swarm Optimization (MDPSO) algorithm, behavioral arbitrators based on behavior trees shall be generated automatically for Khepera IV robots. After an in-depth literature research on the state-of-the-art, a grammar shall be defined compatible with the chosen behavioral arbitrator structure as well as the MDPSO algorithm. Using Webots, a high-fidelity robotics simulator, an arbitrator for a flocking scenario without obstacles shall then be designed using MDPSO and compared against a manually designed solution. If successful, the approach shall then be challenged with a flocking scenario including obstacles.

Recommended type of project:  Semester project / Master project

Work breakdown: 40% theory, 40% software, 20% experimentation

Prerequisites: broad interest in machine learning, control algorithms and robotics, good command of C/C++

Keywords: machine learning, robot control, automatic design and optimization

Contact: Cyrill Baumann

References:

[1] C. Baumann and A. Martinoli, “A Noise-Resistant Mixed-Discrete Particle Swarm Optimization Algorithm for the Automatic Design of Robotic Controllers” to appear in IEEE Congress on Evolutionary Computation (2022)
[2] Duarte M., Sancho Moura O., and Anders Lyhne C., “Evolution of Hybrid Robotic Controllers for Complex Tasks” Journal of Intelligent & Robotic Systems (2015) 78: 463
[3] M. Colledanchise and L. Natale, “Analysis and Exploitation of Synchronized Parallel Executions in Behavior Trees,” IEEE/RSJ International Conference on Intelligent Robots and Systems (2019)

Multi-Robot Gas Distribution Mapping and Source Localization in Simulation

Assigned to Yasmine El Goumi

Gas Distribution Mapping (GDM) and Gas Source Localization (GSL) are useful techniques for real world applications such as environmental monitoring. Implementations so far have focused on 2D mapping and localization [1][2], but a 3D approach would be very valuable, given the tridimensional nature of the plume. Researchers at DISAL are focusing on exploiting quadrotors for gas detection missions in 3D [4] [5]. Moreover, using an array of robotic assets (e.g., sensor nodes, UAVs, wheeled robots), could be beneficial for improving the final gas map.

This project will explore the benefits of a multi-robot architecture in a mission that aims at achieving simultaneous gas distribution mapping and gas source localization. An appropriate multi-robot architecture will be selected and algorithms will be developed to achieve both tasks at the same time. The student will focus on cooperative and collaborative strategies to enhance the performance of the system.

Recommended type of project: semester/master project

Work breakdown: 40% theory, 60% simulation

Prerequisites: Broad interest in robotics, very good programming skill (Python/C/C++), knowledge of ROS, strong interest in research

Keywords: gas source mapping, quadrotors, multi-robot system

Contact: Chiara Ercolani

References:

[1] Lilienthal, Achim, and Tom Duckett. “Creating gas concentration gridmaps with a mobile robot.” Proceedings of the IEEE/RSJ International Conference on Intelligent Robots and Systems Vol. 1. 2003, p. 118-123.

[2] Lilienthal, Achim J., et al. “A statistical approach to gas distribution modelling with mobile robots-the kernel dm+ v algorithm.” Proceedings of the IEEE/RSJ International Conference on Intelligent Robots and Systems, 2009, p. 570-576.

 [3] Ercolani, Chiara and Martinoli, Alcherio “3D Odor Source Localization using a Micro Aerial Vehicle: System Design and Performance Evaluation” Proceedings of the IEEE/RSJ International Conference on Intelligent Robots and Systems, 2020, p. 6194-6200

[4] Ercolani, Chiara, et. al. “Clustering and Informative Path Planning for 3D Gas Distribution Mapping: Algorithms and Performance Evaluation” Robotics and Automation Letters, 2022, p. 5310-5317.

Path Planning for Gas Distribution Mapping

Gas Distribution Mapping (GDM) is a useful technique for real world applications such as environmental monitoring. Implementations so far have focused on 2D mapping [1][2], but a 3D approach would be very valuable, given the tridimensional nature of the plume [3]. Researchers at DISAL are focusing on exploiting quadrotors in gas mapping algorithms to obtain a 3D map of the environment [4]. On top of gas estimation techniques, research focuses on path planning methods that promote quality of the map during a mission. Path planning methods considered so far only estimate the next best sensing location. However, interesting results could be achieved by considering a longer horizon for path planning [5].

The objective of this student project is to research long horizon path planning techniques that could be relevant in a mission with scarce input data, such as a gas mapping mission. The student will then choose one or two promising algorithms and implement them in simulation to assess their performance.

 

Recommended type of project: semester/master project

Work breakdown: 50% theory, 50% implementation

Prerequisites: broad interest in robotics, very good programming skill (Python/C/C++), knowledge of ROS, strong interest in research

Keywords: gas source mapping, quadrotors, path planning

Contact: Chiara Ercolani

References:

[1] Lilienthal, Achim, and Tom Duckett. “Creating gas concentration gridmaps with a mobile robot.” Proceedings of the IEEE/RSJ International Conference on Intelligent Robots and Systems Vol. 1. 2003, p. 118-123.

[2] Lilienthal, Achim J., et al. “A statistical approach to gas distribution modelling with mobile robots-the kernel dm+ v algorithm.” Proceedings of the IEEE/RSJ International Conference on Intelligent Robots and Systems, 2009, p. 570-576.

[3] Bing, Luo, et al. “Three-dimensional gas distribution mapping with a micro-drone.” 34th Chinese Control Conference. IEEE, 2015, p. 6011-6015.

[4] Ercolani, Chiara and Martinoli, Alcherio “3D Odor Source Localization using a Micro Aerial Vehicle: System Design and Performance Evaluation” Proceedings of the IEEE/RSJ International Conference on Intelligent Robots and Systems, 2020, p. 6194-6200

[5] Popović, Marija, et al. “An informative path planning framework for UAV-based terrain monitoring.” Autonomous Robots 44.6 (2020): 889-911.

Gas Source Localization Using a NAV in Cluttered Environments

Deploying robots for Gas Source Localization (GSL) tasks under hazardous scenarios can significantly reduce the risk for human and animals typically engaged in such missions. Thanks to their motion agility and relatively low disturbance on the plume, rotary-winged Nano Aerial Vehicles (NAVs) have become qualified candidates for this task [1]. The chaotic and unpredictable nature of gas dispersion makes the GSL a challenging task. Many existing works have been developed under simplified assumptions, such as constant quasi-laminar airflow that is emitted to an obstacle-free environment, where the plume can be approximated using a well-characterized shape [2]. However, these assumptions do not hold under realistic environments, in particular in presence of obstacles.

This project will deploy a NAV in cluttered environments to a accomplish GSL task, leveraging a 3D bioinspired GSL algorithm in combination with a Deep Computational Fluid Dynamic (CFD) network. Bioinspired algorithms [3] take inspiration from natural living species, such as moths and bacteria, which take wind information as an important indicator to guide their motion for plume tracking. However, the sensing payload limitations and the propeller disturbance make it difficult to deploy commercially available small-scale anemometers on NAVs. In [3], a DeepCFD network is proposed that takes a geometrical map of the environment as input and produces a corresponding wind map as output. The student will focus on the development of a DeepCFD-based bioinspired algorithm enabling the 3D navigation with a NAV in an environment characterized by the presence of one or more simple obstacles. The performance of this algorithm will be compared with an alternative probabilistic algorithm that leverages a DeepCFD network with similar structure for the plume, trained in the very same cluttered environment. The algorithm will be first tested in simulation and, if time allows and results are promising, in a wind tunnel.

Recommended type of project: semester project, master project

Work breakdown: 30% theory, 50% simulation, 20% experiment

Prerequisites: passion on solving issues, broad interest in robotics, good programming skill (Python/C++), knowledge of ROS, experience with Webots will be an asset

Keywords: gas source localization, navigation, deep learning, quadrotors, Webots

Contact: Wanting Jin

References:

[1] J. Burgués, V. Hernández, A. Lilienthal, and S. Marco, “Smelling Nano Aerial Vehicle for Gas Source Localization and Mapping,” Sensors, vol. 19, no. 3, p. 478, Jan. 2019, number: 3.

[2] J. A. Farrell, J. Murlis, X. Long, W. Li, and R. Carde, “Filament-Based Atmospheric Dispersion Model to Achieve Short Time-Scale Structure of Odor Plumes” Defense Technical Information Center, Fort Belvoir, VA, Tech. Rep., Jan. 2002.

[3] C. Ercolani and A. Martinoli, “3D Odor Source Localization using a Micro Aerial Vehicle: System Design and Performance Evaluation,” in 2020 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS). Las Vegas, NV, USA: IEEE, Oct. 2020, pp.6194–6200.

[4] M. D. Ribeiro, A. Rehman, S. Ahmed, and A. Dengel, “DeepCFD: Efficient Steady-State Laminar Flow Approximation with Deep Convolutional Neural Networks,” arXiv:2004.08826 [physics], May 2020, arXiv: 2004.08826.

Sensor Network Development for Gas Source Localization

In environmental monitoring tasks, static sensor networks [1][2] are highly complementary assets compared to mobile platforms, given their higher energetic autonomy and measurement quality. In addition, they benefit from the ability to make synchronous parallel measurements in space, enabling to capture time-variant features of the environment. A new static sensor node for gas source localization tasks has been designed at DISAL. It consists of four modules: a core computational board (M5Stack Core2 [3]), a gas sensing module, a wind sensing module, and a battery pack. To deploy an extendable and manageable sensor network that comprises a sufficient number of sensor nodes in real applications, some additional factors must be considered: first, the energetic autonomy needs to be sufficient for long-run experiments; second, the communication protocol needs to be reliable and extendable.

To solve these problems, a system that controls the wake-up and sleep modes of all sensors and manages the measurements through wireless communication is required. As communication protocol, we have chosen MQTT [4], a lightweight publish/subscribe messaging system designed for setting up bidirectional communication between the host machine and multiple devices with minimal network bandwidth. In order to facilitate a hybrid-mobility sensing system [5] that coordinates the static sensor network with robotic sensing assets, a management and communication system for the sensor network will be designed under the ROS framework. The student will work with real hardware and test system performance in physical gas source localization scenarios.

Recommended type of project: semester project, master project

Work breakdown: 30% theory, 70% development

Prerequisites: passion on solving issues, broad interest in sensor networks, good programming skill (Python/C/ Arduino), knowledge of ROS will be an asset

Keywords: Sensor Network, Arduino, ROS, MQTT

Contact: Wanting Jin

References:

[1] J. Burgués, V. Hernández, A. J. Lilienthal, and S. Marco, “Gas distribution mapping and source localization using a 3D grid of metal oxide semiconductor sensors,” Sensors and Actuators B: Chemical, vol. 304, p. 127309, Feb. 2020.

[2] A. Marjovi, A. Arfire and A. Martinoli, “High Resolution Air Pollution Maps in Urban Environments Using Mobile Sensor Networks,” 2015 International Conference on Distributed Computing in Sensor Systems, 2015, pp. 11-20, doi: 10.1109/DCOSS.2015.32.

[3] https://docs.m5stack.com/

[4] https://mqtt.org/

[5]  W. C. Evans, D. Dias, S. Roelofsen, and A. Martinoli, “Environmental field estimation with hybrid-mobility sensor networks,” in 2016 IEEE International Conference on Robotics and Automation (ICRA). Stockholm: IEEE, May 2016, pp. 5301–5308.

Local navigation for the inspection of steel structures

Community-critical steel infrastructure, such as energy transmission and telecommunication lines, steel bridges, offshore rigs, etc… require regular inspection, in order to guarantee the integrity of the structure and good operation. Current inspection approaches usually require the use of heavy equipment, as well as visual inspection performed by technicians, which exposes them to potentially dangerous situations. Furthermore, inspections usually imply down-time for the structure under inspection, which is costly and inefficient. The main point of interest is therefore to identify fatigue signs within a structure, without requiring to stop its usage. Drones are potentially good candidate to achieve this goal.

It was already assessed in a previous project how a drone can identify key points (typically beam intersections where stress tends to accumulate) within a modeled steel structure. This project should therefore leverage these previous findings and evaluate how a drone can navigate around a designated location to provide good observations within the structure. This will imply to evaluate how the navigation shall be performed (local map representation [1], standard reactive obstacle avoidance, path planning algorithms [2,4], …) and what types of sensors are required to do so successfully (ToF camera, LiDAR, stereo vision, RGB cameras, …). Everything will be performed in the Webots simulator [3]. ROS will likely be used as well. Furthermore, depending on the progress of the project, some experiments could be carried out on real drones if time permits.

Recommended type of project: semester project

Work breakdown: 50% theory, 50% simulation

Prerequisites: Broad interest in robotics, very good programming skills (C/C++, Python). Knowledge in control theory, ROS, Webots, git. Having attended the course “Aerial robotics” or similar is recommended.

Keywords: Webots, simulation, quadrotors, auto-pilot, flight-controller, computer vision, navigation, path planning, ROS.

Contact: Lucas Wälti

References:

[1] Oleynikova, Helen, Zachary Taylor, Marius Fehr, Roland Siegwart, and Juan Nieto. “Voxblox: Incremental 3D Euclidean Signed Distance Fields for on-Board MAV Planning.” In 2017 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), 1366-73, 2017. https://doi.org/10.1109/IROS.2017.8202315.

[2] Oleynikova, Helen, Michael Burri, Zachary Taylor, Juan Nieto, Roland Siegwart, and Enric Galceran. “Continuous-Time Trajectory Optimization for Online UAV Replanning.” In 2016 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), 5332-39, 2016. https://doi.org/10.1109/IROS.2016.7759784.

[3] Webots: https://cyberbotics.com/doc/reference/index

[4] Yang, Liang, Juntong Qi, Jizhong Xiao, and Xia Yong. “A Literature Review of UAV 3D Path Planning.” In Proceeding of the 11th World Congress on Intelligent Control and Automation, 2376-81, 2014. https://doi.org/10.1109/WCICA.2014.7053093.

Point cloud segmentation of infrastructural steel elements

Community-critical steel infrastructure, such as energy transmission and telecommunication lines, steel bridges, offshore rigs, etc… require regular inspection, in order to guarantee the integrity of the structure and good operations. Current inspection approaches usually require the use of heavy equipment, as well as visual inspection performed by technicians, which exposes them to potentially dangerous situations. Furthermore, inspections usually imply down-time for the structure under inspection, which is costly and inefficient. The main interest is therefore to identify fatigue signs within a structure, without requiring to stop its usage. Drones are potentially good candidate to achieve this goal.

This project aims at understanding and modeling online the geometry of steel infrastructures from 3D measurements (typically generated by a stereo camera pair or Time of Flight camera) in real time. The detection can possibly be coupled with computer vision techniques. A suitable representation should be selected (raw point cloud, voxel grid [2], geometry primitives fitting, …). Ways of segmenting the point cloud shall also be explored [3], by possibly leveraging the characteristic of the steel structure. The purpose for the perception of the geometry is to enable autonomous navigation along and within a steel infrastructure. Existing tools such as the Point Cloud Library (PCL [4]) shall be used. It will likely be required to work with ROS, for which previous experience is preferable. Everything will be performed in simulation using the Webots simulator [1].

Recommended type of project: semester project

Work breakdown: 50% theory, 50% simulation

Prerequisites: Broad interest in robotics, very good programming skills (C/C++, Python). Knowledge in control theory, ROS, Webots, git. Having attended the course “Aerial robotics” or similar is recommended.

Keywords: Webots, simulation, quadrotors, auto-pilot, flight controller, point cloud, 3D, segmentation.

Contact: Lucas Wälti

References:

[1] Webots: https://cyberbotics.com/doc/reference/index

[2] Oleynikova, Helen, Zachary Taylor, Marius Fehr, Roland Siegwart, and Juan Nieto. “Voxblox: Incremental 3D Euclidean Signed Distance Fields for on-Board MAV Planning.” In 2017 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), 1366-73, 2017. https://doi.org/10.1109/IROS.2017.8202315.

[3] Nguyen, Anh, and Bac Le. “3D Point Cloud Segmentation: A Survey.” In 2013 6th IEEE Conference on Robotics, Automation and Mechatronics (RAM), 225-30, 2013. https://doi.org/10.1109/RAM.2013.6758588.

[4] Point Cloud Library (PCL): https://pointclouds.org/ 

Design and analysis of modular and scalable Model Predictive Control for MAVs performing formations

The flexibility of control algorithms has a significant effect on the performance outcomes of a distributed robotic system. In this vision, the algorithms should be both modular and scalable so that they should allow subsystems (or robots) to join and leave with minimal supervision effort and the performance should not degrade with the increasing number of subsystems. [1] Model Predictive Control (MPC) is an optimization-based, closed-loop control strategy that has gained significant popularity in the last decade. [2] Distributed and decentralized versions of MPC try to achieve this flexibility; however, the rigid structures of solvers prevent the algorithms from revealing their full potential.

In this project, the main objective is to design and analyze an MPC-based formation control algorithm for Micro Aerial Vehicles (MAVs) which should be both scalable and modular. In other words, any number of robots should be able to enter/exit the swarm performing formation without any supervision and fault/danger, and this number can go up to tens of robots.

This algorithm will be realized in two different simulation environments structured in Matlab and Webots. The initial environments will be provided. The student should, first, perform an in-depth literature review to identify the potential types of MPC’s and solver structures. After deciding on a promising type, the student should prototype this in Matlab and show the effectiveness of the algorithm with respect to the chosen metrics. In the second half of the project, this algorithm will be implemented in Robot Operating System (ROS) and tested in a high-fidelity simulator Webots. A similar performance analysis will be carried out and reported. Further references are listed as follows [3], [4], [5].

Recommended type of project: Semester/master project

Work breakdown: 20% theory, 50% software, 30% experimentation

Prerequisites: Good command of C/C+, Matlab, prior experience with (or willingness to learn) ROS, basic knowledge about MPC and solvers

Keywords: Scalability, MAVs, Distributed/decentralized MPC, PnP, robustness

Contact: Izzet Kagan Erunsal

References:

[1] Raković, S. V., & Levine, W. S. (Eds.). (2018). Handbook of model predictive control. Springer.

[2] Eren, U., Prach, A., Koçer, B. B., Raković, S. V., Kayacan, E., & Açıkmeşe, B. (2017). Model predictive control in aerospace systems: Current state and opportunities. Journal of Guidance, Control, and Dynamics, 40(7), 1541-1566.

[3] Zeilinger, M. N., Pu, Y., Riverso, S., Ferrari-Trecate, G., & Jones, C. N. (2013, December). Plug and play distributed model predictive control based on distributed invariance and optimization. In 52nd IEEE conference on decision and control (pp. 5770-5776). IEEE.

[4] Lindqvist, B., Sopasakis, P., & Nikolakopoulos, G. (2021, April). A Scalable Distributed Collision Avoidance Scheme for Multi-agent UAV systems. In 2021 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) (pp. 9212-9218). IEEE.

[5] Bodenburg, S., Kraus, V., & Lunze, J. (2016, July). A design method for plug-and-play control of large-scale systems with a varying number of subsystems. In 2016 American Control Conference (ACC) (pp. 5314-5321). IEEE.

Spring Semester 2022-2023

The student projects for the next spring semester will be posted at the beginning of November 2022.