{"id":12854,"date":"2025-08-25T08:18:45","date_gmt":"2025-08-25T12:18:45","guid":{"rendered":"https:\/\/blogs.mathworks.com\/student-lounge\/?p=12854"},"modified":"2025-08-25T15:11:37","modified_gmt":"2025-08-25T19:11:37","slug":"designing-a-quieter-future-how-a-student-team-from-fh-aachen-won-the-smart-city-aachen-hackathon","status":"publish","type":"post","link":"https:\/\/blogs.mathworks.com\/student-lounge\/2025\/08\/25\/designing-a-quieter-future-how-a-student-team-from-fh-aachen-won-the-smart-city-aachen-hackathon\/","title":{"rendered":"Designing a Quieter Future: How a Student Team from FH Aachen won the Smart City Aachen Hackathon"},"content":{"rendered":"
\n
<\/div>\n
What if you could redesign your city to sound better? That\u2019s the question a team of Applied Mathematics and Computer Science students from a unique dual study program, jointly offered by FH Aachen, RWTH Aachen and local training companies, set out to answer at the Smart.City Hackathon 2025.<\/span><\/div>\n

A Hackathon for a More Livable City<\/h2>\n
The Smart.City Hackathon 2025<\/span><\/a> in Aachen was a three-day innovation sprint where students, professionals, and mentors collaborated to develop creative solutions for urban challenges. The 2025 theme was: \u201cMake Aachen more livable – with smart ideas!\u201d<\/div>\n
Participants tackled real-world problems in areas like sustainability, mobility, education, culture, and more. For this team, the motivation was clear: \u201cWe wanted to push our boundaries and see how far we could go in just two days.\u201d<\/div>\n
A friend spotted the event and rallied a group of classmates. So many joined that they ended up forming two separate teams! The students from FH Aachen University were very motivated to participate in the event, as the incentive was to contribute to making living in their own city nicer.<\/div>\n

The Challenge: Tackling Urban Noise Pollution<\/h2>\n
The team chose the Freestyle Hack track, which allowed them to define their own problem. The team, who named themselves \u201cNoise Aware,\u201d focused on a problem that affects millions but is rarely seen: urban noise<\/span><\/a> \u2013 an invisible but serious health threat. Aachen is known to be a noisy city<\/span><\/a>, and while static noise maps exist (required by EU law and updated every five years<\/span><\/a>), they\u2019re not interactive and don\u2019t help explore how to improve the situation.<\/div>\n
The team\u2019s idea: simulate a dynamic noise map using real-time sensor data and explore how different materials, like trees, grass, or concrete, affect sound propagation. The goal? Provide guidance for urban planners and engineers designing new public spaces.<\/div>\n

The Solution: Simulating Soundscapes with MATLAB<\/h2>\n
Due to limited access to real-time data, the team used simulated data for their proof of concept. They built a sound simulation in MATLAB, modeling how sound waves reflect and dampen across different surfaces.<\/div>\n
They rapidly prototyped several machine learning models \u2013 k-nearest neighbors, linear regression, and neural networks \u2013 to predict noise levels under varying conditions like time of day and weather.<\/div>\n
\n
\n
%<\/span> normalization<\/span><\/span><\/div>\n<\/div>\n
\n
X = data{:,1:end-1};<\/span><\/div>\n<\/div>\n
\n
mu = mean(X);<\/span><\/div>\n<\/div>\n
\n
sigma = std(X);<\/span><\/div>\n<\/div>\n
\n
X_norm = (X – mu) .\/ sigma;<\/span><\/div>\n<\/div>\n
\n
<\/div>\n<\/div>\n
\n
Y = data.noiseLevel;<\/span><\/div>\n<\/div>\n
\n
<\/div>\n<\/div>\n
\n
% training\/test split<\/span><\/span><\/div>\n<\/div>\n
\n
cv = cvpartition(numSamples, ‘HoldOut’<\/span>, 0.2);<\/span><\/div>\n<\/div>\n
\n
idxTrain = training(cv);<\/span><\/div>\n<\/div>\n
\n
idxTest = test(cv);<\/span><\/div>\n<\/div>\n
\n
<\/div>\n<\/div>\n
\n
XTrain = X_norm(idxTrain,:)’;<\/span><\/div>\n<\/div>\n
\n
YTrain = Y(idxTrain)’;<\/span><\/div>\n<\/div>\n
\n
XTest = X_norm(idxTest,:)’;<\/span><\/div>\n<\/div>\n
\n
YTest = Y(idxTest)’;<\/span><\/div>\n<\/div>\n
\n
<\/div>\n<\/div>\n
\n
% train a neuronal network<\/span><\/span><\/div>\n<\/div>\n
\n
net = fitnet([10 10]); <\/span><\/div>\n<\/div>\n
\n
net.trainParam.showWindow = true;<\/span><\/div>\n<\/div>\n
\n
net.divideParam.trainRatio = 0.8;<\/span><\/div>\n<\/div>\n
\n
net.divideParam.valRatio = 0.2;<\/span><\/div>\n<\/div>\n
\n
net.divideParam.testRatio = 0.0;<\/span><\/div>\n<\/div>\n
\n
net = train(net, XTrain, YTrain);<\/span><\/div>\n<\/div>\n
\n
<\/div>\n<\/div>\n
\n
% evaluate model<\/span><\/span><\/div>\n<\/div>\n
\n
YPred = net(XTest);<\/span><\/div>\n<\/div>\n
\n
mse = mean((YPred – YTest).^2);<\/span><\/div>\n<\/div>\n
\n
fprintf(‘Mean square error: %.2f dB^2\\n’<\/span>, mse);<\/span><\/div>\n<\/div>\n<\/div>\n
<\/div>\n
\n
\n<\/div>\n
\u201cWe created a noise heatmap from the simulated data, which can be overlaid on official city maps. In our interactive app, users can place trees and see how they affect the noise levels. Our idea of the real App: it would suggest optimal placements for noise-reducing elements.\u201d<\/span><\/div>\n
\n
\n<\/div>\n
\"software_demo_smaller.gif\"<\/div>\n

A Winning Concept<\/h2>\n
The team\u2019s project stood out for its technical depth, creativity, and real-world relevance. Their interactive prototype not only impressed the judges, it also won the MATLAB Prize for its innovative use of simulation and machine learning. The team was also honored during the opening ceremony of the Aachen AI Week 2025, where their project was highlighted as an example of student-driven innovation.<\/div>\n
<\/div>\n
\"\"<\/div>\n

<\/h2>\n
\n
\n<\/div>\n
“Leveraging MATLAB’s ready-to-use Machine Learning and physical modelling tools (e.g., for noise reduction) accelerated our workflow and helped us focus on model optimization and data insights.”<\/span><\/div>\n
\n
\n
\n<\/div>\n

Key Takeaways: From Hackathon to Startup<\/h2>\n
For some team members, this was their first hackathon. For others, it was a chance to apply their experience from game jams to a real-world challenge. Beyond the technical learning, they discovered:<\/div>\n
    \n
  • Hackathons are fun<\/span> \u2013 you get to explore open-ended problems that match your interests.<\/li>\n
  • MATLAB\u2019s rich library of built-in tools<\/span>\u2013 like those for acoustics and AI\u2014made it easy to explore ideas quickly. Surprisingly many ready-to-use features are available, helping move fast and stay creative.<\/li>\n
  • There\u2019s startup potential<\/span> \u2013 you get connected with university institutions that support entrepreneurial ideas.<\/li>\n
  • Networking matters<\/span> \u2013 you get to meet mentors and peers, including founders from Startups.<\/li>\n<\/ul>\n

    <\/h2>\n
    <\/div>\n
    <\/div>\n<\/div>\n