NILM 2020 livestream

The NILM 2020 livestream has started:

The full agenda can be found on the NILM website.

NILM 2020 to be held online in November

The NILM 2020 Workshop will be a one day online conference held during the week commencing 16 November 2020. The workshop will be jointly organised by the International and European NILM Workshop teams, and will be free to attend. For the first time, the event will appear as a workshop of the larger ACM BuildSys conference, and accepted papers will be published in the ACM digital library. The paper submission deadline is 21 August, although abstracts should be registered by 14 August. Further details (including the conference date, registration and paper submission) will appear on the NILM 2020 Workshop website as they are finalised.

For context, this workshop represents the culmination of the 5th International NILM Workshop and the 7th European Workshop, as shown by the table below.

I look forward to seeing you online!

DEBS 2020 NILM Grand Challenge

The DEBS Grand Challenge is a series of competitions in which both academics and professionals compete with the goal of building faster and more accurate distributed and event-based system. This year, the DEBS Grand Challenge focuses on NILM, with the goal of detecting when appliances contributing to an aggregated stream of voltage and current readings from a smart meter are switched on or off. The track has already closed to submissions, but the submissions will be presented at the DEBS 2020 conference, which will be held on Thursday 16th July. The whole conference is virtual and registration is free for participants. Registrants have access to the 5 day online program, including tutorials, keynotes, and main conference presentations, as well as access to the Slack discussion channel. The papers can also be downloaded via the online programme.

IDEAL Household Energy Dataset released

The IDEAL Household Energy Dataset was recently announced by researchers at the University of Edinburgh. The data set is under embargo until Autumn 2020. The data set description reads:

The IDEAL Household Energy Dataset comprises data from 255 UK homes. Alongside electric and gas data from each home the corpus contains individual room temperature and humidity readings and temperature readings from the boiler. For 39 of the 255 homes more detailed data is available, including individual electrical appliance use data, and data on individual radiators. Sensor data is augmented by anonymised survey data and metadata including occupant demographics, self-reported energy awareness and attitudes, and building, room and appliance characteristics.

George Hart's 1984 progress report: "Nonintrusive Appliance Load Data Acquisition Method"

Stephen Makonin recently made George Hart's 1984 progress report available, titled "Nonintrusive Appliance Load Data Acquisition Method". This is one of the earliest reports describing a NILM system, and shortly proceeds Hart's 1985 second progress report and 1992 summary report. Happy reading!

On another note, George Hart also gave the keynote presentation at the 2016 International NILM Workshop in Vancouver, where he talked about NILM in the 1980s, and his more recent interests in sculpture and education.

From Hive to Bulb

Last month I said a slightly sad farewell to many great friends as I left Hive. The smart home and internet of things fields present some truly interesting data science challenges, and over the past four years I've been fortunate enough to help people understand their energy usage, detect leaks in their house's plumbing and detect failures in their boiler. It's genuinely been an awesome journey, and I want to say a massive thank you to everyone I worked with along the way.

Earlier this month I started a new role as a senior data scientist at Bulb. Bulb are the UK's fastest growing energy supplier, with a mission to make energy simpler, cheaper and greener. The primary reason I joined Bulb was the opportunity to help their members save energy, and I'm excited to be working back in the field of smart metering. Hopefully this will also be good news for this blog too!

Towards reproducible state-of-the-art energy disaggregation

This week Nipun Batra will present a paper which aims to improve the reproducibility of state-of-the-art NILM at BuildSys 2019 in New York. The full paper is available online, which describes the new experiment and disaggregator APIs in NILMTK, along with a number of algorithms implemented in the nilmtk-contrib repository. The work will also be presented during the demo session to give NILM researchers some hands on experience using NILMTK's new disaggregation algorithms.