Nikan Doosti

Aalto Design Factory

Puumiehenkuja 5, Espoo,

I’m a machine learning researcher, currently working as a research associate focusing on AI-assisted Product Development, mainly integration of data-driven and physics-based deep learning in advanced additive manufacturing (metal 3D printing) processes. This research is conducted within the Engineering department as part of Prof. Jan Akmal project in the Material2Products group research group at Aalto Design Factory, Aalto University.

I have built foundation and experience in DL/ML, Physical-Simulation, Computational Design, and Digital Fabrication; especially enthusiastic about expanding into Drug Design and Healthcare, and in general, applicability of AI in conjunction with science & engineering.

Before this, I worked as a machine learning engineer, building decision-making systems for business optimization. I also had a chance to join the Artificial Intelligence Aided Design and Manufacturing (AIDAM) group at the Max Planck Institute for Informatics (MPI-INF) as an research intern, doing my master’s thesis. During my time there, I focused on Computational Fabrication, specifically, Physics-Informed Deep Learning for Structural Optimization under the supervision of Dr. Vahid Babaei, the AIDAM group leader, and in collaboration with Prof. Julian Panetta, an Assistant Professor at the University of California, Davis. Before this, my work centered around classical digital image processing, deep computer vision, and data science.

Just to state the obvious, I am not doing AI because it’s trendy thing; I was working on ML/DL back in 2017 in my third-year of bachelor’s even though no ML course was part of my curriculum. So yeah, I am truly passionate about it, and I don’t like using it for building TikTok filters even though there’s good amount of money in it. I’m here to solve core problems that people will notice when the downstream consequences and products are being used.

In my spare time, I play video games, read about history, and I like standup comedy. I’m also passionate about open-source projects and consider GitHub my primary social media platform.

I’m always open to collaborate and having a discussion around research, from ideation to dissemination, industry challenges, and even setting up a local server to play Factorio/DoTA2/CS2 and so on! So do not hesitate to contact me!

news

Jun 2024	I am now actively seeking a PhD position where I can apply my passion for research and further contribute to my field.
Jan 2024	I have resigned from my engineering position and took a couple of months to focus on personal growth and career direction, preparing myself for the next step in my professional journey.
Mar 2023	I have been promoted to the position of Team Leader at Panafor
Oct 2022	I defended my master’s thesis with full marks at Iran University of Science and Technology, one of the most prestigious universities in Iran.
Apr 2022	I joined Panafor as a full-time machine learning engineer.

latest posts

Jan 01, 2024	An empty post as the template

selected publications

SCF
Topology Optimization via Frequency Tuning of Neural Design Representations

Nikan Doosti, Julian Panetta, and Vahid Babaei

In Proceedings of the 6th Annual ACM Symposium on Computational Fabrication, Virtual Event, USA, 2021

Abs DOI Bib HTML PDF Code Poster Slides

Structural topology optimization seeks to distribute material throughout a design domain in a way that maximizes a certain performance goal. In this work, we solve the topology optimization problem by parameterizing the designs via recently introduced coordinate-based neural networks. Specifically, we show that networks with Fourier feature mapping can achieve state-of-the-art performance. Our method enables the realization of a range of designs using a single mesh via tuning the frequency content of the solutions independently of the finite element discretization grid. This frequency control offers attractive properties, such as mesh-independent results and sub-pixel filtering that leads to appropriate designs for upsampling. We demonstrate our method on the compliance minimization problem, optimizing for the stiffest possible structure within a weight budget for a prescribed set of loads.
@inproceedings{10.1145/3485114.3485124, author = {Doosti, Nikan and Panetta, Julian and Babaei, Vahid}, title = {Topology Optimization via Frequency Tuning of Neural Design Representations}, year = {2021}, isbn = {9781450390903}, publisher = {Association for Computing Machinery}, address = {New York, NY, USA}, url = {https://doi.org/10.1145/3485114.3485124}, doi = {10.1145/3485114.3485124}, booktitle = {Proceedings of the 6th Annual ACM Symposium on Computational Fabrication}, articleno = {1}, numpages = {9}, keywords = {neural networks, generative design, Topology optimization}, location = {Virtual Event, USA}, series = {SCF '21} }