Case: DPD & Fruitpunch AI
Give data science students and engineers a bunch of data and the task of using this information for greener cities, and they will come back to you with a variety of ideas. This is evident from the interim presentation by engineers from the AI for Good Community by Fruitpunch AI and students from JADS showing their first models and results. DPD Netherlands’ challenge is now entering the next phase. Without anyone having to press a button or a person having to review video footage frame by frame, it is possible to map how green a city or street is. Green as in amount of grass and trees. But also by measuring how many cars, buses, trucks and pedestrians are in a particular place and what their CO2 emissions are. These are the two projects developed using after a call from DPD Netherlands, Fruitpunch AI and JADS. This case is an abstract of the full article as published at Logistiek.nl.
The challenge
The brief was very broad. Develop computer vision models that could be used to make the city greener. It was deliberately not specified that it had to be about sustainable parcel delivery. “That is already a hot topic within DPD itself,” explains innovation specialist Janne Wijnands on behalf of DPD. “We put this challenge away precisely to get a better understanding of the technology as a company. The projects make AI tangible.” Ultimately, DPD wants to discover how the company can contribute to greener cities, in addition to electrifying its fleet. For example, by adding specific sensors or 360-degree cameras to delivery trucks.
The result
Weekly meetings were held to test how feasible the plans were, after which two projects were developed into a proof of concept in 10 weeks. The participating students and engineers chose which project they wanted to join to spend time on – in their spare time.
The first group of students and engineers investigated models to help determine how overgrown a given spot is. On the one hand, the amount of trees and grass, and on the other, the variation in vegetation, based on the number of pixels in a (still) video image.
Project 1
The second project is to produce a model that will make it clear how many cars, buses, trucks and pedestrians are moving around a given place. Being able to automatically recognize the type of vehicle will succeed. What is still proving difficult is precisely determining CO2 emissions. Partly because one truck in a particular view is stationary, while another is driving.
Turning ideation phase into startup phase
JADS also emphatically watched the presentations and proofs of concept of the two projects. If the engineers and students want to take it further, it is possible to set up a startup. Jonie Oostveen of JADS Playground can then help concretise and validate the problem and draw up a business model. This can easily take 12 months. “Whether these projects have a chance of success is hard to say. I hope there are enthusiasts who go ahead with it, so we can help them shape it further.”
We see this as a starting point for creating other models
We see this as a starting point for creating other models
“Our aim is always to make these kinds of ideas open source, so that multiple companies can contribute to – in this case – a greener city. We see this as a starting point for creating other models.”
We are happy to help them further shape the concept
We are happy to help them further shape the concept
“Whether these projects have a chance of success is hard to say. I hope there are enthusiasts who go ahead with it, so we can help them shape it further.”
Why partner up with JADS
JADS provides innovative data science programs at undergraduate, graduate and postgraduate level, conducts cutting edge data science research and offers great business opportunities with a continually growing ecosystem. From startups to SMEs and large corporates, our business partners gain access to talent, high level knowledge and business opportunities.
Partner up with JADS like 300+ other organizations and businesses and shape your data-driven future.