This workshop will bring together interested students, academics and industrial participants from the field of Cyber-Physical Systems. Overall the workshop will discuss how CPS is best described by the bringing together a variety of techniques and concerns. In particular, we will have invited speakers who will talk about human interfaces, speech-based interactions, assistive technologies, and modern perceptual systems including the use of machine learning. 

The lecture is divided into two main parts: (i) an overview of mass market, commercially available robotic vacuum cleaner, the Dyson 360Eye, and (ii) a review of the history of the image based SLAM which provides the 'heart' of the navigation system for the 360Eye.

The 2018 cohort of RAS CDT students will be showcasing the research work being undertaken at Edinburgh Centre for Robotics through a series of demos and interactive sessions at the 2019 Edinburgh International Science Festival.  The students will be based in the Learning Centre Level 2 in the National Museum of Scotland and they will be there from 10.00 - 17.00 hours on 11th - 15th April.

Martin will be presenting:

- Learning to Decompose and Disentangle Representations for Video Prediction. J. Hsieh, et al. (2018)

Abstract:

CST is making people, property and physical data collection safer with autonomous robotics. Mantis, our modular and accessible robotic vehicle collects data autonomously, monitoring and patrolling perimeters reliably with a variety of sensors and AI-driven plug-ins.

For large scale properties this provides reliable, accessible and integrated perimeter security.

The first “International Workshop on Neurorobotics and Neural Disorders (IWNND2019)” is a scientific event and a platform for exchange between researchers, technology developers and clinicians around commercial and/or research application of Neurorobotics in the context of neural disorders, like Parkinson’s and Alzheimer’s disease.

Abstract: Smart devices are becoming increasingly affordable and ubiquitous; whether they are seamlessly embedded in the environment, or carried by people and robots, they continually generate sensor data about the world that surrounds them. Recent advances in machine learning have revolutionised the ability of smart sensors to perceive and interpret context, infer human activities and react to human preferences. This talk will highlight challenges and opportunities in designing machine learning techniques to solve problems of context inference and control.

Talk details and link to meeting schedule below.

Title
Cooperative Deep Multi-Agent Reinforcement Learning

Iordanis Chatzinikolaidis will be presenting:

- Dynamic Locomotion in the MIT Cheetah 3 Through Convex Model-Predictive Control. J. Di Carlo, et al. 

Please note this meeting has been postponed.  New date will be advised asap