🚀 Transforming the Future of Home Robotics at DysonAs a pivotal member of Dyson's Robotics Research team, I spearhead the development of cutting-edge robotic technologies poised to revolutionise home products. In this role, I serve as both a technical leader and hands-on innovator, architecting and orchestrating the creation of intricate integrated systems that seek to push forefront of domestic robotics.My experience and expertise extends across a dynamic spectrum of domains, each critical to the realisation of groundbreaking robotic solutions:👁️‍🗨️ Computer Vision: Unleashing the power of visual intelligence to empower robots with perception and recognition capabilities using both classical and machine learnt approaches.🤖 Robotic Manipulation: Crafting dexterous robots adept at precise manipulation and interaction with the environment.🌐 Navigation: Navigating complex environments with seamless precision through advanced spatial awareness.🤖 Autonomous Control & Decision Making: Empowering robots with autonomous decision-making capabilities for optimal performance and adaptability to get jobs done using Finite State Machines, Behaviour Trees and more recently, LLMs.💬 Natural Language Interfaces: Bridging the gap between humans and robots through intuitive and conversational interactions.🔮 Augmented Reality: Harnessing AR technologies to enchance the ways in which people can interact with robots🤝 Human-Robot Interaction: Shaping intuitive and harmonious interactions between humans and robots to make the robots intentions more transparent and predictable.🌌 Simulation: Leveraging virtual environments to quickly prototype, test and optimize robotic systems, accelerating innovation and reliability.Together, we're pushing boundaries and redefining what's becoming possible in home robotics, with a relentless focus on blending advanced technologies to create intelligent, user-centric robots that seamlessly integrate into everyday life.

The goal of my work is to develop software that allows our robot platforms, Socibot and RoboThespian, to engage in novel and rich social interactions with people (e.g. conversing with the robot, getting information about some- thing or someone). This is a challenging task that involves creating the base technological capabilities required to facilitate social interaction (e.g. sensor fusion, object recognition, saliency maps, speech recognition, dialogue systems), and combining these to create engaging and interactive behaviours.These are then applied to specific usage scenarios (e.g. a robot information kiosk giving information about products, events happening in the local area, space travel) where the robot can project relevant information/content and direct this at users in real time. As the robots are humanoid and people can get very close to them, attention to detail with respect to social cues (and safety) is of the upmost importance.Our software architecture is distributed and asynchronous by nature. Written in Python with components in C++, it is well suited to the application domain of interactive robots. All the software that I write is managed via Mercurial and code is documented during development and auto-generated using Sphinx.

I was working on Non-Linguistic Human-Robot Interaction as part of the EU FP7 ALIZ-E Project (www.aliz-e.org). This project was aimed at understanding how robots can be programmed in order to engaging in long-term human-robot interaction. The project focused specifically upon child-robot interaction within a paediatric/diabetic context where the robot (an Aldebaran Nao) may take up the role of a social peer to children in a paediatric and help them with understanding and better managing their condition, as well as providing companionship. The project ran from 2010 to 2014.Following this I researched technologies behind autonomous robot-enhanced therapy (RET) for children with autism disorders (EU FP7 DREAM project, www.dream2020.eu).

I assisted with the teaching of the MSc module on ”Robotics and Control” with industrial robotic arms. This was a practical module, with both on-line and off-line programming covered. I prepared the course material, exercises, assessments, delivered the teaching and marked the coursework. I also supervised MSc final projects using the Aldebaran Nao robot platform. I provided project management guidance, hands on help with programming the robot. I also was a second assessor for these projects.

Investigating the use of R2D2 like sounds during social child-robot interaction through human based experiments. Research focus was placed on what information can be conveyed through non-linguistic robotic sound, what factors impact their production and perception, and utilizing machine learning techniques to uncover mappings between acoustic parameters of the sounds and their affective interpretations, pushing toward automated generation of non-linguistic sounds yielding a particular desired affective interpretation.This research was conducted as part of the EU FP7 ALIZ-E project (www.aliz-e.org) exploring the science and technology behind long-term sustainable child-robot interaction in paediatric settings.