Contact Duncan Wilson if you are interested in these projects.
CASA have been working with Moorfields Eye Hospital to explore the potential for the use of pop-up eye clinics in retail outlets (Brent Cross) to catch up on COVID appointment backlog. Time motion data (x,y data ever 10 seconds) from approximately 4000 patient journies and 4 different activity layouts are available. This project would explore different ways of analysing trends in patient throughput over time and space. Requires data visualisation to inform communication of trends and or data analysis to extract meaning from the data.
[2023]
Problem / Hypothesis: Alongside the Institute of Archeology at UCL we are working with volunteers at House Mill (https://housemill.org.uk) on the River Lea, a Grade 1 listed 18th century tidal mill. Their challenge is the Mill regularly floods causing damage to the wooden structure of the building. Question: Would real-time continuous flood data help inform decision makers on approaches to alleviate the flooding. Approach: Through the deployment of Lora based flood sensors and the installation of timeseries cameras to capture the ebb and flow of the water levels, a reporting tool will be developed to demonstrate data over at least a two month period over spring / summer 2023.
[2023]
In Connected Environments we have a large number of IoT sensors and gateways deployed in the building, around the park and across London. This project will create a visual representation that is part art work / part digital dashboard to help staff and students understand the current status of devices. A secondary aim of the project is to create a reporting tool to allow for the better management of those services.
[2023]
CASA have a number of bat monitors in the Queen Elizabeth Olympic Park that have been counting bat calls since 2017. In this research project we are interested in developing a v2 of the Bat Monitor. The project would focus on implementing a new device using AudioMoth as a sensor plus Raspberry Pi as host for deep learning models to classify bat calls. Knowledge of Edge Impulse and or TensorFlow Lite required. Whilst initial focus is on bats the device could also be used to detect bird calls (using Birdnet model).
[2022]
CASA have a number of bat monitors in the Queen Elizabeth Olympic Park that have been counting bat calls since 2017. In this research project we are interested in exploring how to visualise changing calls volumes given environmental conditions (climate, events, seasons).
Dataset: available - approx 8 bat monitors (bat counts and probability of bat type), 1 weather station (usual weather parameters), event information (concerts, events etc)
[2022]
We have a 4m diameter geodesic dome with 1500 LED’s – how could this be used to create an interactive installation that tells a story about what is happening on the 1PS roof garden or as a building interface on the 3rd floor terrace.
[2023]
Accurately detecting the presence of people while maintaining privacy of identity are the potential benefits of using mmWave sensor technology. This project will explore the development of a novel modular sensor unit that can be implemented to support understanding of space usage.
[2023]
Utilising 3D models of the Olympic Park area and the new UCL East buildings the goal of this project is to build a rig that would allow people to experience the sense of flying through our environment in VR. The focus of the project is to explore how movement of the body can be used to create a sense of floating when combined with the visual input of VR.
[2023]
This research project explores how IoT can be used as a tool for improving water quality and safety in rivers, with a particular focus on continuous monitoring using low cost connected sensors. To achieve this objective, an IoT platform will be deployed that includes sensors installed along the waterway which are able to monitor temperature, oxygen levels and other relevant parameters in real-time.
[2023]
Can Computer Vision help us understand how workshop spaces get used? The project would involve building a prototype camera that implements AI / ML techniques to anonymously track people in Bartlett workshops and visually represent activity trends over time.
Dataset: open source image datasets would be used for training - project would create new datasets through inference on workshop usage. We have Oak 1 and Oak D cameras available for implementation.
[2022]
Project to develop a machine vision system to use object detection algorithms to anonymously count room occupancy in lecture theatres / offices. We have Oak 1 and Oak D cameras available for implementation.
[2022]
Water height in the waterways around the Queen Elizabeth Olympic Park provide risk indicators for managing assets in the area such as the community group managing the House Mill (https://goo.gl/maps/iH62ydGmeu5ZHdWm6). This project will utilise the Lora network deployed in the park to capture water height data over several months. The project will focus on deployment of IoT sensors plus gathering, processing and visualisation of data to help inform decision makers.
[2023]
CASA have several IoT energy monitors on appliances in CASA measuring the coffee machine, kettle, water cooler, printers, screens etc. We are interested in analysis of this data from different perspectives: could time series analysis of this data help inform patterns of behaviour or help predict future activity? Could novel forms of visualisation help us see the data from a different perspective?
Dataset: 8x energy monitors capturing power usage every 2 secs since December 2019.
[2022]
CASA have developed a walkshop web app called Smart Walks which has a database of geolocated technology in the city. The goal of this project would be to explore how this screen based application could be extended to incorporate novel forms of in-situ interaction and presentation of data in the urban environment.
[2022]
CASA are working with the UCL East project team to develop an IoT Strategy to release the large volumes of data generated by modern buildings. Motivated by the technological concept of a smart campus but aware of the socio political risks, we are interested in developing some speculative futures to help frame how future scenarios of this campus of the future could play out. We are keen that the research explores potential futures through physical prototypes, design fictions or works of art.
Dataset: documents on the vision for UCL East / planned programmes / data registers; sample datasets from existing building systems to build proof of concepts
[2022]
The Alice Camera’s mission is to empower the next generation of content creators who have been let down by the lack of innovation and usability of traditional digital cameras. Photogram surveyed over 1,000 content creators and found that 89% wanted a device with the image quality of an interchangeable lens camera but with the connectivity of a smartphone.
By taking advantage of a new opportunity to bring together mirrorless camera technology, micro four thirds lenses and smartphone technology with the advances in computational photography and artificial intelligence, the team at Photogram are drastically rethinking the camera as a computer with an open source ecosystem making it easier to shoot, edit and share content seamlessly whilst being affordable and maintaining a professional optical system.
Project Brief: We are using Google's edge TPU in our camera so edge computing projects using deep learning and TFlite would be good. AR is also something we haven't fully explored. We don't have any preference on the specific technical application so we'd be open to suggestions.
[2022]
To estimate how much photovoltaic energy will be produced, we first need to know where the solar PV systems are. At the moment this is not well known in the UK for all but the largest solar PV systems - particularly domestic systems are poorly documented. This project would develop a methods for citizen scientists to monitor solar potential in their locale.
[2022]
https://activethings.app https://twitter.com/doactivethings/
We are working with Scott Cain (Honorary Senior Research Fellow) Founder and CEO of Active Things. “Active Things is your Active Travel Assistant, helping you find and access (secure) bike parking, routes and more.”
The goal of this project is to identify and map the potential locations in London where cyclists may want to stop and secure their bike for 30 mins to 2 hours. With cycle theft on a continued upwards spiral there is a need for more secure urban parking for bikes. Active Things provide remotely accessible secure bike storage which you pay for by the minute. This is not aimed at storage at home or in the office (which are increasingly well catered for) but the intermediate journeys cyclists would like to make on their bikes – e.g. stopping at the pub for an hour on the way home, calling in at the gym, stopping to pick up some shopping etc. Research by Active Things indicates that risk of theft puts off many cyclists making these intermediate journeys and has a negative impact on cycle use (ie they don’t use the bike if they plan to go to the pub.
Research question: where should the first 1000 lockers be installed to best cater to these intermediate journeys? What kind of profile of use could be imagined in terms of daily usage patterns? Do the current deployments of bike racks match the need for secure parking?
Active Things are also interested in prototype smart lock systems that could be used with their existing cycle storage systems.
Data: Active Things are currently the provider of the TfL bike mapping service. They have expressed intent to share research findings from surveys interviews and feedback from users over the past year.
[2022]