Design Fleet Map Data Collection
A self-driving car, or autonomous vehicle (AV), leans on a high-definition (HD) map: lane lines, curbs, crosswalks, signs and traffic lights, placed to within centimeters. The car compares what its sensors see against that map to work out exactly where it is, which is called localization. When a road changes, the map is wrong until someone drives the road again with sensors and rebuilds that piece. That's the job of this system. A fleet of cars with cameras and lidar drives the roads that need it. The system decides which roads those are, plans the driving, pulls the recordings off the cars when they return, checks the data and hands the good parts to map building. It sounds like logistics, and partly it is. The hard parts are a freshness model for hundreds of thousands of road segments, a terabyte of data per car per day over links that are mostly offline, and knowing when a road really counts as covered.
WaymoPremium lesson · System design round
The Waymo design, worked end to end.
- The API and data modelSystem design · Frequently asked
- The design, drawn out1 diagram
- 5 deep divesWith the tradeoffs
$49 a year · all 107 companies · instant access
20,000+ candidate reports · guide updated
Waymo interview guideKnow what Waymo asks before you walk in.
27 coding · 4 system design · 15 behavioral questions
Get their questions first.
InterviewLegend is independent and not affiliated with or endorsed by any company it covers.