09-16, 10:50–11:20 (Europe/Amsterdam), Bar
In the Netherlands a large share of energy is used by industry. By measuring the energy usage of individual machines in real time it is possible to pinpoint when machines are operating inefficiently and help factories take measures to reduce energy waste. It turns out that in most factories, the biggest source of energy waste comes from idling machines. To be able to give valuable insights and provide relevant alerts to our customers, we set up a machine learning system for standby detection with a “human in the loop”. In this talk we will go over the considerations that go into setting up a machine learning system with a human in the loop and showcase our approach to the problem. No background knowledge is required for this talk.
In the Netherlands a large share of energy is used by industry (>40% compared to only 14% used by households*). Eliminating energy waste in this sector is a big step forward towards a greener future. Therefore, Sensorfact made it its mission to eliminate all industrial energy waste. By measuring the energy usage (electricity or gas) of individual machines in real time it is possible to pinpoint when machines are operating inefficiently and help factories take measures to reduce energy waste.
It turns out that in most factories, the biggest source of energy waste comes from forgetting to turn off machines when they are not used. Flagging idling machines based on their electricity usage may seem like a trivial problem at first, however the large variety in machines and production processes makes this a lot harder than you would expect. To be able to give valuable insights on idling machines and provide relevant alerts to our customers, we set up a machine learning system with a “human in the loop”.
In many settings it is perfectly fine to embed a machine learning model in a process without any human interference. However, there are cases where it is better to keep a human in the loop. The most obvious use cases are those where there is simply no room for error, for example in medical applications. However, also in less life threatening it can be beneficial to have a human act as gatekeeper ensuring high quality outputs. In this talk we will go over the considerations that go into setting up a machine learning system with a human in the loop and showcase our approach to the problem, using the case of standby detection. We will share learnings from our own experience and along the way give you an overview of the (open source) tools we chose to use for the different facets of the project.
No background knowledge is required for this talk. If you are looking for inspiration on how to build a machine learning system with a human in the loop or if you are curious about sustainability use cases this talk may be interesting for you.
*https://www.clo.nl/indicatoren/nl0052-energieverbruik-per-sector
No previous knowledge expected
Lieke is lead data scientist at Sensorfact, a company aiming to eliminate all industrial energy waste for SME’s. In her role she focusses on the data fueled products that help their consultants to efficiently and effectively give advice to customers. Before joining Sensorfact she worked as a data science consultant at Vantage AI and completed a PhD in econometrics.