Tune into the 2018 NILM Workshop live:

Monday 1st October – Tuesday 2nd October 2018

Fraunhofer-inHaus-Center, Duisburg, Germany

About the NILM Workshop


The EU NILM Workshop is now in its fifth year! The aim of this workshop is to bring together researchers that are working on the topic of energy disaggregation in both academia and industry. The workshop will feature two full days of conference-style material, following a similar format to the EPRI EU NILM 2017 conference. We hope to livestream the presentations via youtube, and also make videos of presentations available afterwards.

Please check out our “Call for Speakers & Demos” to find out what you can contribute!

Who should attend?

If you are an individual or an organisation interested in energy disaggregation then this is the workshop for you. The workshop is an open forum to encourage collaboration within the field and we welcome all levels of knowledge in this area, even if you are relatively new to NILM. In case you can’t join us in person, we also hope to live-stream the presentations via YouTube to allow for remote viewing. However, we do encourage you to attend so you benefit from the networking opportunities, which offer real value and the ability to meet fellow researchers in the industry.

What is NILM?

NILM stands for Non Intrusive Load Monitoring. This is the process which disaggregates your total energy consumption into the energy readings of an appliance over a period of time, allowing users to see how much power a particular appliance has used during that time.


Monday 1st October 2018

09:00 Registration & coffee
10:00 Welcome and introduction
– Oliver Parson – Workshop introduction Watch video I View presentation 
– Prof. Dr.-Ing. Holger Vogt – Welcome to Fraunhofer inHaus-Center Watch video I View presentation
10:05 Opening keynote 
10:30 Comfort break
11:00 Session 1: Innovative applications. Session chair: Oliver Parson 
– Peter Davies (Green Running) – Disaggregation/Blockchain P2P Energy Trading Watch video I View presentation
– Benoit Leprettre (International Electrotechnical Commission) – Ongoing standardization work on sensing devices for NILM View presentation
– Cristobal Lowery (Baringa Partners) – Predicting residential occupancy and hot water usage from high frequency, multi-vector utility data Watch video I View presentation
12:00 Lunch break
13:30 Session 2: Vendor talks. Session chair: Peter Davies
– Stephen Galsworthy (Quby) – Toon energy waste checker: Appliance level diagnostics at scale View presentation
– Filip Milojkovic (Homepulse / EWUS) – NILM-as-a-service for heating systems: a business case View presentation
– Antoine Clais (Voltaware)
– Erol Can Ün (SmartB) – Learnings from Industrial NILM View presentation
14:50 Coffee
15:20 Session 3: Data collection. Session chair: Nikolaus Starzacher 
– Hader Aguiar Dias Azzini (University of Campinas) – A NILM Research and Development Project In Brazil View presentation
– Stephen Makonin (Simon Fraser University) – Data, Datasets, and Data Engineering View presentation
– Grant Fisher (Pecan Street) – Dataport View presentation
16:20 Lightning talks – Oliver Parson View presentations
16:30 Poster & demo session
18:00 Workshop group photo
Evening social event at Buergerhof

Tuesday 2nd October 2018

09:00 Coffee & networking
10:00 Session 3: Algorithms presentation session. Session chair: Timo Bernard
– Henning Lange (Carnegie Mellon University) – Overcoming the challenges of Variational Inference in the context of NILM Watch video I View presentation 
– Marina Dorokhova (EPFL PV-LAB) – Unsupervised algorithm for disaggregating low-sampling rate electricity consumption of households Watch video I View presentation 
10:40 Break
11:00 Session 4: Deep learning in NILM. Session chair: Filip Milojkovic
– David Murray (University of Strathclyde) – Transferable low rate NILM using deep neural network architectures Watch video I View presentation 
– Shubham Bansal (Discovergy GmbH) – DeepNilm meets Bayes: Estimating uncertainty of deep learning models for non-intrusive load monitoring Watch video I View presentation 
– Patrick Huber (Lucerne University of Applied Sciences and Art, iHomeLab) – NILM @ Industry Watch video I View presentation 
12:00 Lunch break. Includes an optional guided tour through inHaus-Center.  
13:30 Keynote Talk:
– Ulrich Greveler (Rhein-Waal University of Applied Sciences) – Consumers fear smart metering and load monitoring – who is to blame?
14:20 Session 5: Evaluating NILM algorithms presentation session. Session chair: Shubham Bansal 
– Shamim Ahmed (Fludia) – Evaluation of ML algorithms on NILM using semi-synthetic training dataset Watch video I View presentation 
– Lucas Pereira (M-ITI/LARSYS and prsma.com) – SoCoL: Semi-automatic Collaborative Labelling for NILM datasets Watch video I View presentation 
15:00 Break
15:30 Discussion session. Session chair: Nikolaus for the collaboration session, all organisers for plans for next year.
– Nikolaus Starzacher (Discovergy GmbH) – Developing the SuperMeter: Call for collaboration Watch video I View presentation 
– Open discussion
– Plans for next year
16:30 Workshop close


Please note: The agenda is subject to change. 

NILM Workshop - Task

Register your place 


Monday 1st – Tuesday 2nd October 2018


9.00am – 5pm


Fraunhofer-inHaus Center
Forsthausweg 1
D-47057 Duisburg


A limited number of free tickets are available until 31st August on a first-come-first-served basis.


Travel & Accommodation

Local Travel

Düsseldorf Airport is the closest to the venue, and the closest train station is Duisburg Central Station. If you are flying into Düsseldorf airport, take the train to Duisburg Central Station. The following trains are available: S1 (takes 10 minutes), RE2, RE5 & RE6 (takes 25 minutes). You will then need to take a bus or taxi to the venue. Download our travel document to view all options.

Getting Around

There is parking for 20 cars at the venue and plenty of additional parking is available at the main university building 100 metres away. Local taxis are available. Uber does not operate in the area.

Hotel Information

The venue has kindly recommended a list of hotels and accommodation options. Please download to view and contact them directly to book.


Oliver Parsons

Oliver Parson

Centrica Hive Limited

Peter Davies NILM Expert

Peter Davies

Green Running

Niki Davies

Niki Davies

Green Running

Niki Davies

Nikolaus Starzacher


Niki Davies

Timo Bernard

Fraunhofer IMS

Call for Sponsors

We have a number of sponsorship options for the workshop, including opportunities to give a full presentation, present a demo, exposure on the website and slides, and also to sponsor an evening reception. Please contact us for further information.


We would like to thank our sponsors for their support and enabling us to continue running the NILM Workshops.

Green Running Logo


Verv (a Green Running brand) is a smart home hub using AI to provide advanced insights into individual appliance usage and condition so users can save money on their energy bills, reduce their carbon footprint and keep their home safe.

British Gas Logo

British Gas Connected Homes

Control. Comfort, convenience and cost savings. This is what smart energy products and services in a connected home can give our customers.

British Gas Logo


Discovergy aim to provide full transparency and control over energy consumption and production.

British Gas Logo


The Fraunhofer Society is a German research organisation with 69 institutes spread throughout Germany, each focusing on different fields of applied science.

Green Running Logo


smartB is reinventing energy management for commercial buildings. The cloud-based solution reveals savings potential through real-time monitoring and the automatic analysis of usage patterns.

Green Running Logo


Smart Meter intelligence for profitable energy analytics.

Contact Us

If you have any questions or suggestions about the upcoming NILM Workshop please complete the form and a member of the team will get back to you as soon as possible.

You can also follow our latest updates on Twitter our join our LinkedIn group to connect with other NILM enthusiasts.

10 + 11 =

Join our NILM LinkedIn Group