← Waymo Interview Insights

Waymo·Machine Learning Engineer·Technical Phone Screen·Senior

Senior
May 2026

Summary

Waymo ML engineer interview with a physics/spatial reasoning problem involving two colliding objects and a diagram. Pretty niche question, not your typical ML theory stuff.

Questions Asked (1)

Q1

Given two objects on a collision course, use the provided diagram to determine what the outcome of the collision will be.

Technical Trade-offsAdaptability & Ambiguity
Author's notes

This one threw me a bit.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Start by clarifying the assumptions and constraints of the scenario, then walk through a structured reasoning process that combines physics-based prediction with ML model considerations. Emphasize safety and uncertainty handling, and conclude with a clear outcome and potential mitigations.

Pro tip: Show that you think beyond the immediate collision: discuss how you would validate the prediction with simulation and real-world data, and how you'd handle edge cases. This demonstrates a safety-first mindset valued at Waymo.

1. Clarify the Scenario

Ask clarifying questions about the objects (e.g., vehicles, pedestrians), their speeds, masses, and the environment. Confirm that the diagram is to scale and that no other factors (e.g., weather, road conditions) are at play.

2. Identify Relevant Physics and ML Models

Discuss the physical principles (momentum, energy) that govern the collision outcome. Then, explain how ML models (e.g., trajectory prediction, collision detection) would be used to predict the outcome given sensor data.

3. Analyze the Diagram and Predict Outcome

Use the diagram to estimate trajectories and impact point. Apply physics to determine post-collision velocities or deformation, and consider uncertainties in the prediction.

4. Discuss Safety and Decision-Making

Explain how the system would use this prediction to make decisions (e.g., braking, evasive maneuvers) and prioritize safety. Mention redundancy and fail-safes.

5. Validate and Iterate

Describe how you would validate the prediction using simulation and real-world data, and how you would handle edge cases or model uncertainties.

Key Points to Mention

  • Conservation of momentum and energy in collision analysis
  • Trajectory prediction using ML models (e.g., LSTM, transformer-based)
  • Uncertainty quantification and probabilistic outcomes
  • Sensor fusion (LiDAR, radar, cameras) for accurate object detection and tracking
  • Safety fallback mechanisms and redundancy in autonomous systems
  • Simulation and scenario-based testing for validation

AI-generated suggestions, not part of the candidate's original notes. May be inaccurate — verify before relying on them.