Aims

The aims of the course are to:

• To provide the fundamentals of end-to-end heterogeneous communication system design including testing and characterisation through conventional, i.e., RF, and unconventional, i.e., molecular and acoustic, communication techniques.
• To provide hands-on experience on channel characterisation and implementation of modulation and detection techniques.
• To develop methods for optimising communication links through iteration of design equations based on empirical data.
• To gain an appreciation of the importance of interoperability and unconventional communication methods in the forthcoming ICT landscape characterised by heterogeneous networks, e.g., Internet of Everything, beyond 5G/6G networks.

Content

The objective of extending our control and connectivity to underexplored environments, e.g., underwater, intra-body, with an ever-increasing resolution has led to the emergence of new unconventional communication modalities, such as acoustic communications and molecular communications (MC), which uses molecules as information carriers. MC has emerged as a bio-inspired, low-complexity and energy-efficient technique to interconnect heterogeneous artificial nanoscale networks and biological networks within the framework of Internet of Everything (IoE). Some of the applications expected to be enabled by MC include continuous health monitoring and smart drug delivery with intrabody Internet of Bio-Nano Things (IoBNT), and programmable biological systems, e.g., human gut-brain axis, plant-animal communication networks. At macroscale, it is expected to complement electromagnetic wireless communication systems in 6G and beyond-6G networks through orthogonal links in applications, such as covert communications. Considering that the smell is nothing more than our perception of volatile molecular signals, MC is also expected to enable interesting applications regarding virtual reality.

The recently introduced IoE framework is positioned to exploit the heterogeneity of current and next-generation communication and networking technologies (both conventional, e.g., RF, and unconventional, e.g., acoustic and molecular) to extend our connectivity to the entire universe, which is itself a natural IoE, an inherently heterogeneous network of everything we perceive. The countless number of opportunities that can be enabled by IoE through a blend of heterogeneous ICT technologies across different scales and environments and a seamless interface with the natural IoE impose several fundamental challenges, such as interoperability, ubiquitous connectivity, energy efficiency, and miniaturization.

In this project, the students will work in groups of three to build a one-way heterogeneous communication link between a human controller and a mini drone in an indoor environment. The end-to-end link involves an acoustic link, an MC link, and an RF link. This multi-modal link, which is built upon the concatenation of three different communication technologies, embodies one of the fundamental properties of the IoE framework, i.e., the interoperability of heterogeneous communication networks and technologies. Accordingly, the communications in this multi-modal link is realised through the following steps:

1. The human controller, who wishes to control the vertical movement of the mini drone utters ‘up’ or ‘down’ commands to indicate the direction of the drone movement.
2. The acoustic signal is then converted to digital signals in a PC through its microphone, which processes the signal for wake word detection, distinguishing ‘up’ and ‘down’ commands from silent and background noise, using an available speech recognition model
3. The binary command signals, i.e., ‘up’ and ‘down’, are then encoded into different durations of MC pulse signals, which are generated by the release of volatile isopropyl alcohol (IPA) molecules into air (MC channel) by an electrically controllable sprayer at a constant release rate.
4. The propagated airborne molecular signals are then detected by a commercial gas sensor, i.e., MQ-3, which generates an electrical potential at its output with the voltage proportional to the concentration of the IPA molecules adsorbed on the sensor surface.
5. A microcontroller (i.e., Arduino Uno) receiving analog input from the MQ-3 sensor is used to decode the transmitted binary molecular messages.
6. The controller of an RF-controlled mini drone, which is rewired to connect to the receiver microcontroller, receives from the microcontroller the up and down command signals that are recovered by a decoding algorithm.
7. The controller is then used to transmit RF control signals to the mini drone via the RF channel, which are translated to the up or down movement of the mini drone.

During the project, particular focus will be devoted to the airborne MC link. Once the MC link has been implemented, the students will be expected to experimentally characterise the airborne MC channel in terms of channel response, inter-symbol interference (ISI), and noise, by also drawing on the theoretical MC channel models available in the literature. Major challenges in construction of the MC link will be ISI due to dispersion of MC signals, time-varying channel properties due to stochastic propagation of molecules and potentially varying airflow profiles, and nonlinearity of the channel response due to the saturation of the receiver. 

Week 1

Implementation and tests of the bidirectional Arduino UNO — PC (MATLAB) serial connection, MC Transmitter, and MC Receiver

Week 2

Characterisation of the airborne MC channel based on empirical data collected through MC signal transmission tests with varying communication parameters (e.g., transmitter-receiver distance, pulse durations, transmission rate). Comparison with theoretical MC channel models available in the literature.

Implementation of acoustic link by implementing a wake word detection algorithm for ‘up’ and ‘down’ commands using an available speech recognition model.

Week 3

Implementation of the RF by rewiring the controller of a mini drone in connection with the receiver microcontroller through a relay circuit.

Integration of End-to-End Acoustic/Molecular/RF Communication Channel

Week 4

Performance tests and optimisation of the multi-modal communication link.

MINI LECTURES

Two mini lectures will be delivered to:

1. Introduce Internet of Everything and Molecular Communications.
2. Review the literature on channel characterisation, transmitter & receiver designs, modulation & detection techniques for MC.

Aims

The aims of the course are to:

• Learn advanced topics of robotics (underactuated robotics, robot learning, soft robotics, human robot interactions, and distributed robotics)
• Fundamentals (theories and methodologies) of advanced robotics researches
• Practical implementation of advanced robotics technologies

Objectives

As specific objectives, by the end of the course students should be able to:

• Extend the knowledge of introductory robotics to more advanced ones to carry out research
• Learn research techniques and skills for robotics projects
• Work effectively with collaborators in robotics projects
• Deliver professional presentations and communication of robotics projects

Content

This course aims to extend the knowledge and skills of students in designing and developing autonomous machines and researching robotics-related topics. Beyond the Introduction to Robotics course given in MT, the Advanced Roboticscourse will focus on more advanced topics such as Robot Learning, Underactuated Robot Control, Soft Robotics, Human-Robot Interaction, and Multi-Agent Systems, which are not covered in the introductory course.   

Lectures (2 lectures per week, a total of 16 lectures):

1. Introduction (2L; F Iida, F Forni)

1. 1. Course overview; 
   2. History and landscape of robotics; 
   3. basic knowledge and theories (kinematics, dynamics, planning/search); 
2. Underactuated Robotics (4L; F Forni)
   
   1. Problem formulation and modelling
   2. Control approaches of underactuated systems
   3. Case studies
3. Robot Learning and Adaptation(2L; R Antonova)
   
   1. Model-based learning approaches
   2. Model-free learning approaches 
   3. Optimization methods and case studies
4. Soft Robotics (2L; F Iida)
   
   1. Soft material/body robot modelling; 
   2. Soft actuators and sensors; 
   3. Control and learning of soft robots; 
5. Human-Robot Interaction 1 (2L; F Iida)
   
   1. Introduction to human-robot interaction
   2. Theoretical frameworks (spatial, nonverbal, verbal interactions)
   3. Research methods, applications, robots in society
6. Distributed Robotics, Multi-Agent Systems (2L; F Iida)
   
   1. Planning and control in multi-robot systems
   2. Methods for learning coordination and cooperation in multi-agent systems
7. Coursework Presentations (F Iida, F Forni)

Aims

The aims of the course are to:

• Learn advanced topics of robotics (underactuated robotics, robot learning, soft robotics, human robot interactions, and distributed robotics)
• Fundamentals (theories and methodologies) of advanced robotics researches
• Practical implementation of advanced robotics technologies

Objectives

As specific objectives, by the end of the course students should be able to:

• Extend the knowledge of introductory robotics to more advanced ones to carry out research
• Learn research techniques and skills for robotics projects
• Work effectively with collaborators in robotics projects
• Deliver professional presentations and communication of robotics projects

Content

This course aims to extend the knowledge and skills of students in designing and developing autonomous machines and researching robotics-related topics. Beyond the Introduction to Robotics course given in MT, the Advanced Roboticscourse will focus on more advanced topics such as Robot Learning, Underactuated Robot Control, Soft Robotics, Human-Robot Interaction, and Multi-Agent Systems, which are not covered in the introductory course.   

Lectures (2 lectures per week, a total of 16 lectures):

1. Introduction (2L; F Iida, F Forni)

1. 1. Course overview; 
   2. History and landscape of robotics; 
   3. basic knowledge and theories (kinematics, dynamics, planning/search); 
2. Underactuated Robotics (4L; F Forni)
   
   1. Problem formulation and modelling
   2. Control approaches of underactuated systems
   3. Case studies
3. Robot Learning and Adaptation(2L; F Iida)
   
   1. Model-based learning approaches
   2. Model-free learning approaches 
   3. Optimization methods and case studies
4. Soft Robotics (2L; F Iida)
   
   1. Soft material/body robot modelling; 
   2. Soft actuators and sensors; 
   3. Control and learning of soft robots; 
5. Human-Robot Interaction 1 (2L; F Iida)
   
   1. Introduction to human-robot interaction
   2. Theoretical frameworks (spatial, nonverbal, verbal interactions)
   3. Research methods, applications, robots in society
6. Distributed Robotics, Multi-Agent Systems (2L; A Prorok)
   
   1. Planning and control in multi-robot systems
   2. Methods for learning coordination and cooperation in multi-agent systems
7. Coursework Presentations (F Iida, F Forni)