Safety1st

safety first!
landing
unity simulation
high risk score + badge + confidence
low risk score + badge + confidence
simulation history

Studies show that women are 17–73% more likely than men to be seriously injured in a crash of equal severity — 17% more likely to be killed, and up to 73% more likely to be hurt.

One major reason: almost all crash testing still centers on the Hybrid III 50th-percentile male dummy.

As for the so-called female model, the Hybrid III 5th-percentile female? It's actually just a scaled down version of the regular male dummy, not a realistic representation of women’s anatomy or biomechanics. This mismatch means seatbelts, airbags, and crumple structures don't offer women the same protections they do to men. In fact, in some cases, these safety features can even increase injury risk for women, in particular pregnant women, as malfitted airbags and uncompromising seatbelts result in worse crash outcomes.

Overall, women are 2–3x more likely to suffer moderate injury at the same crash energy, while car manufacturers continue to ingore this fact.

Safety1st, safety first

In short, we built Safety1st. Because needless to say, safety comes first.

Safety1st is a pre-production crash design tool that helps car manufacturers test how safe their vehicles are for everyone, including women, pregnant or not, before they are built.

Through a web app, users enter basic vehicle details such as mass, crumple zone length, airbags, and cabin rigidity, and then choose an occupant profile, including presets for female and pregnant passengers who are often overlooked in crash testing.

Safety1st runs a simplified Unity crash simulation that applies basic physics calculations to estimate forces and displacement on the vehicle and dummy. It takes the data inputted from the user to generate real models, not AI guesses. But beyond just the math, here's the key: an explicit focus on sex-specific differences, taking into account additional risk factors for women such as faster body acceleration and closer positioning to the wheel and airbag. We aggregate these factors together with calculations of industry standard metrics such as thorax deflection and femur load into one comprehensive baseline risk score. That risk score is then passed to a machine learning model and web scraper to return a more polished score, a confidence level, and a short explanation of which design choices increase injury risk.

By visualizing risk early in the design process, Safety1st allows engineers to compare configurations and make safer, more inclusive decisions before cars ever reach the road.

Safety features, featuring safety!

🕹 unity crash simulator—an interactive WebGL crash mockup with adjustable crash side, dummy size, posture, and vehicle parameters.

⚙️ baseline physics engine—based on real world crash data computing industry standard metrics while incorporating sex-specific differences to show how design choices affect occupant injury risk.

🤖 gemini-powered scoring—Gemini 2.5 Flash computes a risk score, confidence level, and factor breakdown tied to the entered parameters.

🚗 dynamic vehicle + dummy builder presets for female, male, pregnant, and size variations with auto-filled anthropometric baselines.

💽 mongodb data layer persistent storage with async fetch for saved simulations and crash history.

🎛 accessible ui modes high-contrast, colour-blind-safe, dyslexia-friendly, reduced-motion, and large-text themes for inclusive testing.

Accomplishments that we're proud of

We learned how to blend three worlds that never usually touch: physics-based crash simulation, machine-learned injury prediction, and real-time frontend interactivity.

More importantly, we built a project that puts women and pregnant occupants back into crash testing, a group often overlooked by current safety standards. It feels good to build something that helps close that gap and moves us one small step closer to making cars safer for everyone 🚗 (◕‿◕｡)

What's next for Safety1st

🕹️ crash simulation lab Unity WebGL scenes showing frontal and side impacts with adjustable speed, cabin strength, and crash geometry.

📂 scenario history save crash runs, replay them, and compare different designs to spot safety wins and tradeoffs.

🧩 mix-and-match crash labs explore wild what-ifs like lighter cars with stronger cabins or bulky SUVs with tiny crumple zones and see who walks away.

🌍 a world where safety fits all not a niche feature list, but a quiet revolution toward cars built for real humans, not textbook placeholders.

Citations!

Bose, Dipan, et al. “Vulnerability of Female Drivers Involved in Motor Vehicle Crashes: An Analysis of US Population at Risk.” American Journal of Public Health, U.S. National Library of Medicine, Dec. 2011, pmc.ncbi.nlm.nih.gov/articles/PMC3222446/#:~:text=Tolerance%20to%20traumatic%20injury%20may,positioning%20of%20the%20head%20restraint.

“Fatality Facts 2023: Males and Females.” IIHS, www.iihs.org/research-areas/fatality-statistics/detail/males-and-females.

“Ratings.” NHTSA, www.nhtsa.gov/ratings.

Ready, Vernon. “Disproportionate Danger: Women Face a Higher Risk of Injury in Car Crashes.” Ready Law, 7 Nov. 2023, www.ready-law.com/disproportionate-danger-women-face-a-higher-risk-of-injury-in-car-crashes/.

Samarrai, Fariss. “Study: New Cars Are Safer, but Women Most Likely to Suffer Injury.” UVA Today Archives, 25 July 2019, archive.news.virginia.edu/content/study-new-cars-are-safer-women-most-likely-suffer-injury/.