• wireless sensor network
  • citizen science
  • smartphone sensing
  • vector control
Mosquito-borne disease has a major impact on human health, income and mortality in over 100 countries affecting over half the world’s population. Malaria alone is transmitted by over 40 species of mosquito, caused 627,000 deaths worldwide in 2012 and reduced African GDP by over $12bn. Control programmes rely heavily on mapping and modelling tools which are limited in that they are static in time and space. We will embed acoustic monitoring sensors in handheld devices such as smartphones and wristbands, which can detect (unique, species-specific) wing-beats from mosquitoes and alert the carrier to their presence. Detection data will be available for real-time analysis by researchers, who will combine it with ultra-high resolution (30m pixel) remote imaging to determine environmental factors such as vegetation composition and structure and distance to water bodies. This multi-layered data will support the design of better-targeted, more effective vector control programmes.

[ gimme more ]

  • animal welfare
  • disease control

A collaboration between the Department of Zoology and Engineering Science at the University of Oxford, OpticFlock monitors the welfare of farm animals by analysing movement of animals in a pen.

I'm involved primarily as a data scientist, analysing and collecting data of chicken flocks. I also manage the project's processing infrastructure, including the data sharing and storage facilities, processing servers and web servers. The project is lead by Prof Marian Dawkins and Prof Stephen Roberts.

[ gimme more ]

  • smartphone sensing
  • energy usage
  • android

I'm collaborating with Dr Philipp Grünewald, of the The Environmental Change Institute at the University of Oxford, to build a smartphone app that helps people report how they use energy in their house. The app is part of a national project, called METER and led by Phil, to understand what people use electricity for. Find out more about on the project's website.

My involvement is in the design and development of a user-facing app that will act as a survey. Prospective participants will be asked to sign up for a trial, which will involve all people in their household. When they do so, everyone living in their house will receive a small, cheap smartphone and report what activity they're doing throughout the day, with a particular focus on their electricity consumption.

I'm extremely excited to work with Phil on this, and can't wait to receive a kit myself. Keep up to date with the progress on the project's blog pages.

[ gimme more ]

  • citizen science
  • crowdsourcing
  • biodiversity monitoring
  • smartphone sensing
The New Forest Cicada (Cicadetta montana s. str.) is the only cicada native to the UK. During May to July it sings with a very characteristic high-pitched song, which is at the limits of human hearing, and is particularly difficult for most adults to hear. Sightings of the cicada within the New Forest date back to 1812, but the last unconfirmed sighting was in 2000. However, it's quite likely that colonies remain undiscovered in less visited parts of the forest. The New Forest Cicada Project aims to equip the millions of visitors to the forest with a smart phone app that can detect and recognise the song of the cicada, and hopes to rediscover it in 2013.

[ gimme more ]

  • informal transport
  • mobile payments
  • cape town

Cape Town's public transit authorities are in need for data to understand how to improve quality of service and increase the accountability of transport operators. CiTiVAN proposes a phone based rating, payment and positioning system that promotes transparency and accountability from the bottom up, giving people a voice about the service they receive.

For more information click here. Note that I have now left the project, which is still on-going. For up-to-date info please visit senseable.mit.edu.

[ gimme more ]

  • air quality monitoring
  • pollution
  • noise levels

From the project's press release, January 2014: Over the past weeks, MIT researchers and LAAB Studio have deployed a system of human-borne sensors to assess and compare air pollution in Hong Kong and Shenzhen. Results are presented as a dynamic map called 'One Country, Two Lungs', currently on show at the Hong Kong / Shenzhen Bi-City Biennale of Urbanism and Architecture (until the end of February 2014). The installation highlights an asymmetric exposure to pollution between the inhabitants of Hong Kong and Shenzhen, showing how air quality across one of the world’s largest metropolitan areas changes step by step. One Country Two Lungs also provides a glimpse into how new applications of sensor and tracking technology can be used for monitoring air pollution in cities on a personal, human scale.

Read more on the One Country, Two Lungs website.

[ gimme more ]

  • wireless sensor network
  • energy harvesting
  • msp430
Wireless Sensor Networks often harvest energy from the environment in order to reduce battery replacement and potentially achieve perpetual operation. However, the energy sources used, such as solar and wind power, are variable and their availability can highly influence the lifetime of the nodes. In particular, by knowing in advance what the available resources are, it becomes possible to tune sensing patterns and communication algorithms, coordinating the consumption with the intake. This project models a network of nodes in order to discover how and in what way the above issues can be addressed. To do this, weather forecast information from the web are combined with local observations, collected through a custom-made deployment of TI MSP430 nodes with a miniature wind turbine.
  • energy harvesting
  • single-path routing
  • information throughput
Intelligent energy management is a key challenge in Wireless Sensor Networks. The choice of an appropriate routing algorithm constitutes a critical factor, especially in unstructured networks where, due to their dynamic nature, a reactive routing protocol is necessary. Such networks often favour packet flooding to fulfil this need. One such algorithm is IDEALS, a technique proposed in the literature, which balances energy consumed with information delivered. This paper evaluates the use of a single-path solution with IDEALS to increase efficiency. Simulation results comparing the two approaches show that the single-path algorithm outperforms flooding in terms of energy consumption for any network size. Furthermore the benefit of IDEALS is preserved as its combination with the single-path algorithm maximises information throughput.

Publications

Loading publications is taking a long time. Try reloading the page.