Author: Mark Szepieniec (Data Scientist)

What do fast-moving consumer goods (detergent, shampoo, toiletries, etc.) have in common with energy analytics? This question was among the topics explored at the consumer goods hackathon hosted at Procter & Gamble’s InQbet Campus Brussels Innovation Center last month.

Representing ONZO, I worked on a team exploring how wireless sensor technology and sophisticated data analytics could enable subscription models for goods such as laundry detergent and dishwasher tablets. This could help the producers establish a direct-to-consumer relationship, as well as allow them to target millennials, who tend to have very low brand loyalty, but who are more likely than other groups to sign up to online subscription services.

Where does ONZO’s expertise come in? The answer is that the goal of our team’s concept was to achieve auto-replenishment, i.e. delivering a new package of product just as the current one is running out. To do that, the idea was to capture the machine’s electricity consumption and use ONZO’s analytics to obtain an estimate of the number (and perhaps also the type) of cycles run, which would be an obvious help in determining when to dispatch a new shipment of product to the consumer.

We investigated the potential of the Sigfox network (an IoT-optimized communications network which doesn’t use WiFi or GSM networks and has a range of up to 60km) to transfer enough data from a smart plug to make cycle counting feasible, and concluded that the data rate (which is very low in order to optimise energy consumption) is sufficient for the purpose of home appliance signalling. In addition, we were provided with access to a consumer panel to help validate our concept, and the feedback was very positive.

Thanks to Procter & Gamble for inviting ONZO to the event — it was a very interesting and enlightening experience, and we’re now eager to run a pilot program in this space.


Illustration: Dr. Oriol Camacho