
The Radar Dilemma That Taught a Car to Ignore
In the mid-2010s, autonomous vehicle developers faced a critical engineering tradeoff: how to filter out background radar clutter without blinding cars to stationary hazards. This episode explores the technical logic behind why early self-driving systems were tuned to discount unmoving objects, and how that design choice manifested in real-world autonomy failures.
Listen in the Fylom app.
Radar systems often ignore stationary objects to prevent dangerous phantom braking from soda cans or metal signs.
Engineers face a tradeoff between smooth driving and detecting critical hazards like parked emergency vehicles.
Early autonomous systems struggled to distinguish between a stopped car and permanent roadside scenery like guardrails.
Action suppression logic prevents erratic braking but can lead to fatal failures in object classification.
The twenty eighteen Tempe crash occurred partly because factory emergency braking was disabled to avoid false positives.
Multi-object tracking uses temporal filtering to suppress sensor returns from vegetation and power lines.
- 01Intro1 min
- 02The Sensor Tradeoff Matrix3 min
- 03The Evolution of Filtering Logic3 min
- 04How Machines Learn to Ignore3 min
- 05The Limits of Suppression in Tempe2 min
- 06Outro1 min
- Case 7: Mobileye - SweetStudy
- Computer Vision for Autonomous Vehicles:
- [PDF] Sensor and object recognition technologies for self-driving cars
- Self-Driving Cars That Recognize Free Space Can Better ...
- Recent Advances in 3D Object Detection for Self-Driving ...
- Inside Waymo's Secret World for Training Self-Driving Cars
- Coherent lidar for ride-hailing autonomous vehicles - arXiv
- HWY18MH010.aspx
- Obstacle detection using range-difference events
- Uber in fatal crash had safety flaws say US investigators
- Preliminary Report
- Toward open benchmark tests for automotive lidars, year 1: static range error, accuracy, and precision
- Uber Fatal Self-Driving Crash NTSB Report
- JOURNAL OF LAW, TECHNOLOGY, & THE INTERNET • VOLUME 15 • ISSUE 1
Fylom generates episodes like this on any topic you're curious about.
Fylom episodes are researched, written, and voiced by AI. Automated checks help catch inaccuracies, but episodes aren't reviewed by a human and AI can still get things wrong. Treat them as a starting point, not a source of record — more in our accuracy disclaimer.