Smart De-Icing pilot with Quantillion

Pilot in Delft


How can data help to keep our roads safe and efficient in winter?


  • Project: Smart De-Icing
  • Location: Delft, the Netherlands, with 103.000 inhabitants
  • Challenge: To optimise de-icing policy in Delft by using open data sources
  • Solution: An interactive dashboard that makes use of different open data sources in order to allow the city of Delft to base their de-icing policy on calculated data


Delft was looking to boost the cycling flow in the city during Winter. In order to do so, they decided to improve the process of de-icing by using the city’s open data. The process used to be done mostly manual, relying on the experience and guessing of the people involved. With the rapid change of our climate, experience was no longer enough, and technology was required to optimise the city’s resources. The first goal is thus to formalise the de-icing decision-making process. 



On the basis of the above-described challenge, the company Quantillion has delivered a proof of concept that uses three years of historical data to set up and showcase the workflow of autonomous and predictive de-icing recommendations. Thus, making the process automated instead of manually.This has been presented in an interactive dashboard that shows the calculation of the final recommendation (and data upon which it was based), and shows the assignment mechanism (i.e. resources to routes). With regards to the decision-making capabilities, these were tested gradually for different decision impacts (1. De-ice, yes or no? 2. If so, when? 3. If so, where precisely?) to allow for controlled testing and an accurate insight into the added value of the model.



Results & learnings from this pilot on de-icing:

  • Workflow differs per municipality, and any final solution should leave some room for this
  • There are some legal impairments for having this solution work fully autonomously (i.e. governments care-taking responsibility needs to be ‘checked’ by a political appointee)
  • Involving de-icing companies as end users would improve the provided solution; many municipalities validate the number of dispatches but leave de-icing operations to (sub)contractors
  • The idea and look & feel of the web-app are fresh, intuitive and help with creating new added value in the workflow



For any city that deals with de-icing, the results from this pilot provides a useful example on how innovation and automated decision-making could be realised.  The full implementation of this pilot is in the process of getting implemented in hardware solutions of providers that deliver de-icing solutions to many other cities throughout Europe.



For Quantillion, the next step is to explore the dashboard’s feasibility in a pilot with a number of different municipalities and to start cooperating with ‘operational’ end users (i.e. contractors and subcontractors). For the city of Delft, an important next step is to find out if and how the tender process could be adjusted as a means to ensure that (sub)contractors innovate their working process by using some of the results that were delivered by this pilot.