Ongoing Projects

Development of a soft highly robust robot

Julien Couyoupetrou (Microengineering)

Soft robotics is revolutionizing the field of robotics, allowing to develop robots highly robust and safe to use in human environments. However, existing soft robots have usually locomotion performances still far from those of their natural counterparts or require tethering to an external actuation system or power supply. A possible strategy to design soft robots - currently pursued in our lab - is to use heterogeneous soft deformable modules using off the shelf electronic components that can be assembled into the robot morphology. In this proposed semester project, we seek to expand the current available kit of soft modules developing a new actuated module allowing to complete the basic kit to design a fully functional untethered soft robot. At first, the student will design and fabricate the additional actuated module on an available design concept. He will use lightweight materials, 3D printing technologies and off the shelf electronic components. Secondly, the available design of the complete first basic kit of modules will be used to design and build a proof of concept simple untethered soft robot able to locomote. Finally, the performances of the robot will be assessed in terms of locomotion speed, operational time and robustness.

Type: Semester project
Period: 25.02.2019 - 24.06.2019
Section(s): Microengineering
Type of work: 10% theory, 80% hardware (mechanical and electronics), 10% software
Requirements: CAD (Inventor, SolidWorks or similar), good understanding of mechanisms and materials, Arduino programming and 3D printing experience a plus
Subject(s): soft robotics, bio-inspired robotics, integrated systems
Responsible(s): Davide Zappetti, Olexandr Gudozhnik

Video-audio communication for delivery drones

Leonardo Cencetti (Robotics)

At the Laboratory of Intelligent Systems (LIS) we are developing drones for last-cm delivery. These delivery drones are fully autonomous and can be monitored in real-time with the help of a web-application framework named Dronistics. In order to facilitate operation of the drone and communication between a sender and recipient, a camera, speaker and microphone will be deployed on the drone. The goal of this project is to design the audio and video communication between the sender and the recipient of the drone to ensure the efficiency. The first goal of this project is to perform a comparative study of protocols and existing open-source solutions that can be used to implement bi-directional audio communication and uni-directional video communication. On the sender side, the video and audio should be streamed on a web interface, and on the drone, the algorithm should be running on a Linux-based external computer that is running no GUI (e.g. a Raspberry Pi). The second goal is to design system architecture, make an analysis of the network throughput requirements, audio-video quality, scalability, security and privacy issues, and reliability. The third goal of the project is to implement the outcome of the above work in a real delivery drone and test it in a real scenario. The sender should initiate/join the communication from a web-based interface and the person next to the drone should initiate the communication by clicking a button on the drone.

Type: Semester project
Period: 19.02.2019 - 30.06.2019
Section(s): Robotics Microengineering
Type of work: 30% software architecture, 50% software development, 20% testing
Requirements: Solid understanding of software architectures, network protocols, and web-related technologies
Subject(s): Software Architecture, IoT, audio-video communication
Responsible(s): Anand Bhaskaran, Alessandro Crespi, Przemyslaw Kornatowski
URL: Click here

Long Range communication (LoRa) for Drones

Camille Conrad Aussems (Robotics)

At the Laboratory of Intelligent Systems (LIS) at École Polytechnique Fédérale de Lausanne (EPFL) we are developing drones for last-cm delivery. These delivery drones are fully autonomous with the help of a web-application framework of Dronistics. Reliable long distance Communication in mobile systems (like in the case of Delivery Drone) has always been a challenge that attracts various companies and researchers. LoRa is one of the promising technologies that offer long-range low-power ad-hoc communications. The objective of this project is the implementation of a, LoRa system in the delivery drones (developed at LIS) and the characterization of the communication channel. The first goal of this project is to perform a theoretical understanding of LoRaWAN Networks. This should be followed by a comparative study on various LoRa hardware modules that could be used in drone communication. Finally, various characterization tests for communication channel (for signal strength, bandwidth etc.. ) should be performed with the chosen module(s). The second goal of the project is to implement a LoRa communication between a base station and a real drone developed at LIS. The student can use various open hardware and open software resources for the implementation. The final goal of this project is to perform the feasibility study to extend the implementation securely to use a public network like Swisscom LPM, ThingsNetwork, etc..

Type: Semester project
Period: 19.02.2019 - 30.06.2019
Section(s): Robotics Microengineering School of computer and communication sciences
Type of work: 40% hardware, 20% software, 20% theory, 20%testing
Requirements: Good Understanding of Embedded Systems and Communication Systems
Subject(s): IoT, Communication Systems
Responsible(s): Anand Bhaskaran, Fabrizio Schiano, Przemyslaw Kornatowski
URL: Click here

Precise localization of a LoRa node through a UAV embedded with LoRa 5G

Victor Pierre Guy Delafontaine (MT)

Unmanned aerial vehicles (UAVs) are incredibly versatile tools capable of completing a wide range of applications. Soon the majority of drones will be connected to the cloud for BLOS missions and also perform fully autonomous flights. This will be achieved by using the novel Low Power Networks (LPN) which are characterized by a longer range, lower bandwidth, and low battery consumption. In this project, we aim at improving the localization function of the LoRa LPN network in order to find a node with a precision of less than 20m. This can be applied to scenarios of search and rescue (e.g., missing person in case of an avalanche). Towards this goal, a LoRa gateway will be installed on the drone and a node, to be localized, will be placed at an unknown location. When the node starts beaconing, the position is computed by triangulation, and the drone will plan his trajectory towards this position. Once the UAV is in the vicinity of the node, different algorithms will be implemented and tested for determining the position of the node accurately. The drone will also be embedded with 5G technology (e.g., NB-IOT) to transmit the LoRa node signal to an APN hosted on a Swisscom server. On this server, triangulation algorithms will be run to improve the first position estimation. A simulation environment will be set up (ideally in Matlab and Simulink) to test the software infrastructure before going on the real drone. Hardware in the loop simulations (connecting LoRa gateways on a laptop) could be envisioned to test the whole framework (e.g., LoRa gateway acquisition, communication with the Swisscom server). If time permits, the experiment will be extended to multiple UAVs in order to perform a local triangulation (without the need of running the localization algorithms on the server). Ideally, experiments will be performed in city areas, where the LPN signal is strong, as well as mountain areas where the signal is nonexistent. Tasks of the student: • Setup a simulation environment (preferably in Matlab/Simulink) which will allow to test the whole infrastructure without flying the drone • Install a LoRa Gateway on a drone • Connect the drone to the 5G Network (through NB-IOT technology) • Setup a server which gives the initial triangulation algorithm • Program the flight-path algorithm for reaching the desired area • Program the node-search algorithm based on the RSSI values • Perform flight tests and evaluate the results Ideal Candidate: • Passionate about the IoT and new technologies • Drone experience will be much appreciated • Good programming skills in C, C++, Python and Matlab/Simulink. Any other programming language is appreciated • Familiarity with the networking techniques • Above average academic results • Research oriented personality with hands-on experience The master thesis project will last 6 months and the working place will be in the Swisscom Digital Lab, in the EPFL campus. The master thesis, depending on the quality of the results, should result in a scientific publication. For more details regarding this subject, please consult: [1] "Unmanned Aerial Vehicle Based Wireless Sensor Network for Marine?Coastal Environment Monitoring", Carlos A. Trasviña?Moreno et al., Sensors vol 17, issue 3 [2] "Understanding Autonomous Drone Maneuverability for Internet of Things Applications", Azade Fotouhi et al., WoWMoM 2017 [3] https://lora?

Type: Master project
Period: 04.02.2019 - 02.08.2019
Section(s): Robotics
Type of work: 20%Theory +40%Software +30%+hardware +10%misc
Requirements: control+theory +programming+(Matlab +python+or+similar)
Subject(s): Localization +control+and+estimation
Responsible(s): Fabrizio Schiano, Anand Bhaskaran, Alexandru+Rusu+(Swisscom)

Simulation of multiple Crazyflies quadcopters

Mahdi Nobar (Mechanical Engineering)

At the Laboratory of Intelligent Systems, we develop algorithms for coordinating the navigation of multiple quadcopters. The goal of this project is to develop the infrastructure for testing the swarming algorithms through software-in-the-loop simulation and evaluate their behavior. A preliminary phase involves the modeling of a Crazyflie drone. The result is a file of parameters in a standard format accepted by Gazebo, a dynamics simulator for robotics. Lots of studies about parameters identification and modeling of a Crazyflie exist on the internet. You can adapt them for your purpose. The first phase of the project consists of adding the Crazyflies firmware in the loop. The swarming algorithms are currently implemented in Matlab and Python. With the help of Matlab, Simulink and Python the student will send the input commands to the drones through ROS and develop a simple and intuitive analysis tool to monitor the state of the drones in real-time. The second step will involve the design of a simulation environment in Gazebo (so-called world) which resembles the DroneDome at LIS. Furthermore, the addition of obstacles will enable you to test the swarm in a cluttered environment. To make the obstacle detection possible onboard, the student will have to model a multi-ranger deck in Gazebo and implement a state-of-art obstacle detection strategy. Finally, to investigate the feasibility of embedding the swarm controller in the Crazyflie for a distributed approach, the last stage will imply the exploration of automatic code generation provided by Matlab and Simulink. A Crazyflie will be available for hardware tests. Depending on the quality of the results, the semester project could result in a publication. Previous experience with the cited software is required.

Type: Semester project
Period: 21.02.2019 - 05.07.2019
Section(s): Robotics Microengineering Mechanical Engineering School of computer and communication sciences
Type of work: 20% theory, 50% software, 30% testing
Requirements: Modelling and programming skills (Matlab, Simulink, Python), previous familiarity with ROS.
Subject(s): Swarm robotics, drone formations, simulation
Responsible(s): Enrica Soria, Fabrizio Schiano