AI used for Food Waste monitoring
AI and Sustainable Hospitality: Boon or Bane?
Sustainability in Hospitality — Viewpoint by Willy Legrand
My experience with the usage of AI is related to "Intelligent Scales" used to monitor food waste in the kitchens. This technology is combining scales and a camera that can analyse the content of the bin and help to make a choice when it comes to categorize the type of food waste, for example: trimmings, french fries, ...
This technology seems revolutionary, providing a huge support to the teams to monitor the food waste as all the collected data are used for daily, weekly, monthly reports.
After visiting many different restaurants and kitchens using this technology, it came to my mind that maybe the persons who decided to implement AI technology have been a bit too quick in their decisions. AI is appealing, it seems to solve all our problems but in this particular case there are questions that should have been solved first:
1. is the level of analysis that this technology can provide adapted to the operations?
2. does the person in charge of this technology (most probably the Chef executive) have the time to control and run the machine?
3. is the equipment in capacity of monitoring huge amount of food waste in the rush hours?
4. should all the different kinds of waste be considered food waste? (for example, the watermelon trimmings)
5. when the reports received by the management are not looking good for the Chefs, what strategies are implemented to make them look better?
All these questions are coming from discussions with the Chefs and observations during on site audits.
I fear that AI in this particular matter has been seen as a short cut and nice toy to solve an issue that is much bigger and necessitate a slow and deep approach. Food Waste monitoring is difficult, AI monitoring scales can be a great tool to implement once the teams have already a good experience in monitoring and have been through all the potential questions that may be linked to the monitoring.
So before jumping on a this very advanced solutions, let's implement the traditional systems, let's test them and make sure they are working. Then AI will provide the next level of information that is expected.