I'm a passionate researcher working for autonomous road safety. My research aims at safer traffic for all modes of road users utilising automated technologies, which include but not limited to automated vehicles, decentralised traffic monitoring, and data-driven improvement of transportation systems.

I have a mixed education background in Operations Research and Management Information Systems; self-taught knowledge in Data Science and Machine Learning; as well as solution-driven skills in Artificial Intelligence during addressing theoretical and practical problems. My scientific interests are generally in collective patterns emerged in individual interactions, of which the curiousity also motivated me to pursue a PhD degree. I am approaching the end of my journey at TU Delft and have submitted my PhD Thesis titled "Proactive Collision Risk Quantification in Multi-directional Traffic Interactions". My doctoral research is focused on safety quantification of road user interactions. Most of my papers published during PhD are openly accessible thanks to the TU Delft Library. For every paper, code for experiemnts and instructions to reproduce the results are open-sourced here on GitHub. Below is a more detailed list, where we

• measured two-dimensional spacing between road users [code] [pdf],
• assessed bias-induced explanations for shorter time gaps when human drivers follow automated vehicles [code] [pdf],
• proposed the first unified theoretical framework to probabilistic traffic conflict detection [code] [pdf],
• are proposing (under single-blind peer review) structure-preserving contrastive learning for geospatial time series to learn representations that facilitate downstream tasks [code][pdf],
• are developing (under single-blind peer review) a scalable, context-aware, and generalisable approach to learn the collision risk in traffic interactions from naturalistic driving [code][pdf].

I enjoy thinking, reading, and writing (in both machine and human language), although unfavourably my mind gets overloaded sometimes. I actively post work-relevant updates on LinkedIn and other thoughts as my WeChat Moments.

My digital CV is available at https://yiru-jiao.github.io/cv -- which is very bibliometrically formed for HR's or quantitative eyes. But please, if in any way possible, use a qualitative view to look at me and my research. I believe all researchers in academia would ask the same. We work for a better world, not for producing publications.

Other interesting repositories I have made public include:

• Documented Knowledge Sharing makes open access to the documented knowledge created or summarised by me and my collaborators;
• SSMsOnPlane shares vectorised algorithms to calculate various surrogate safety measures (SSMs), or in another way called, surrogate measures of safety (SMoS) for pairs of road users on an abstracted plane of road, i.e., in a two-dimensional space;
• Two-Dimensional-Time-To-Collision allows for fast computation of two-dimensional adaption of the longitudinal SSMs including TTC, DRAC, and MTTC;
• BirdsEyeTrajectoryReconstructionSHRP2NDS provides information and guidelines to use the reconstructed trajectories of naturalistic crashes and near-crashes in the SHRP2 NDS.
• Reconstruct100CarNDSData reconstructs bird's eye view trajectories of vehicles involved in crashes and near-crashes from 100-Car Naturalistic Driving Study (NDS) radar data;
• MakingRandomBingoCards generates randomised bingo cards (for party games) with custom cells with LaTeX and Python.