Stress Testing Tesla FSD V12 is Getting Difficult...
TLDRIn this video, the host takes the Tesla Full Self-Driving (FSD) system for a rigorous test drive through the narrow and challenging streets of the Berkeley Hills. Despite the system's progress from being uncomfortable to use to now managing most of the drive with occasional disengagements, the host remains impressed with the undeniable progress made in recent months. The FSD system demonstrates smooth handling, cautious behavior around pedestrians, and improved visualizations. However, there are still moments of odd behavior and stress, particularly with speed limit recognition and navigating road closures. The host anticipates further improvements in upcoming software versions, especially in handling construction zones and road closures.
Takeaways
- ๐ FSD V12 has made significant progress in handling challenging driving scenarios, such as the narrow streets of the Berkeley Hills, compared to previous versions.
- ๐ FSD's autopilot visualizations have improved, rendering surrounding vehicles more steadily and accurately than before.
- ๐ FSD exhibits cautious behavior, especially around pedestrians and when approaching parked cars with brake lights on, which can be seen as good defensive driving.
- ๐ฆ There were instances where FSD did not use turn signals when it was the only vehicle on the road, possibly a quirk from being trained on human driving data.
- ๐ ฟ๏ธ FSD had difficulty with a sharp right turn and pulling away from the curb, indicating areas where it can still improve.
- ๐ง FSD struggled with road closures and construction zones, which is a consistent challenge that requires further updates for better handling.
- ๐ก FSD's handling of speed limits was problematic when it adhered to a faded 5 mph sign, causing it to drive unreasonably slow on subsequent roads.
- ๐ฃ๏ธ In narrow sections of the road, FSD demonstrated smooth and natural driving behavior, hugging the road similar to a human driver.
- ๐ง FSD's interaction with road construction workers and adherence to their signals need improvement for better navigation through construction zones.
- ๐ฆ FSD showed good decision-making in an unprotected left turn, waiting for oncoming traffic and ensuring a safe turn.
- โฐ The drive was comfortable and less stressful compared to previous experiences, indicating that FSD is approaching a level of competency where interventions are minimal.
Q & A
What has been the progression of Tesla's Full Self-Driving (FSD) system in the past few years?
-The FSD system has progressed from a system that the narrator wouldn't feel comfortable using in challenging areas like the Berkeley Hills, to being able to manage the majority of the driving with occasional sketchy behavior, and now to a point where the progress is undeniable, despite not being perfect.
What is the role of local news channels in shaping public perception of full self-driving technology?
-Local news channels are criticized for painting full self-driving technology as dangerous, often focusing on sensationalism and misinformation rather than presenting factual information about the technology's capabilities and limitations.
How does Ground News aim to help navigate the challenges presented by biased or misleading news coverage?
-Ground News provides unbiased news coverage, presenting diverse perspectives on global stories. It offers insights into the credibility of news sources, their political biases, and funding sources, allowing readers to form more informed opinions.
What improvements have been made to the autopilot visualizations in the latest versions of Tesla's FSD?
-The latest versions of FSD have significantly improved the autopilot visualizations, rendering cars more steadily and accurately. The visualizations now better represent the actual position and movement of surrounding vehicles, reducing the previous jumpiness.
How does the FSD system handle speed bumps and stop signs in the Berkeley Hills?
-The FSD system does a good job at slowing down for speed bumps, even as low as 1 mph, and recognizing faded stop signs on the ground. It also exhibits smooth driving behavior over neck speed bumps and yields to pedestrians and other vehicles appropriately.
What is the significance of the occupancy network visualizations in the context of FSD?
-The occupancy network visualizations, which have replaced the cone visualizations in version 12, provide a more accurate representation of objects on the road. They are expected to potentially take over the majority of FSD visualizations in the future, offering a more detailed and precise rendering of the environment.
What challenges did the FSD system face during the drive through the Berkeley Hills?
-The FSD system faced challenges such as navigating narrow streets, reacting to road closures and construction zones, and dealing with unclear or missing road markings. It also had instances where it did not use turn signals and struggled with pulling away from the curb smoothly.
What is the narrator's opinion on the current state of Tesla's FSD system after the drive?
-The narrator acknowledges that while the FSD system is not perfect, it has improved significantly, especially in challenging areas like the Berkeley Hills. The drive was comfortable and the system felt competent for the majority of the time, indicating that FSD is progressing towards a higher level of autonomy.
How did the FSD system handle a situation where it encountered a 5 mph speed limit sign?
-The FSD system recognized the 5 mph speed limit sign, even though it was faded and odd-looking, and reduced its speed accordingly. However, it had difficulty adjusting its speed when the limit clearly did not apply, such as on wider roads, leading to the need for manual intervention.
What is the narrator's expectation for future updates of Tesla's FSD system?
-The narrator expects that future updates, particularly versions 12.4 and 12.5, will bring improvements to handling construction zones and road closures. They also hope for better map data for speed limits and smoother interactions with hand-held stop signs in construction areas.
How did the FSD system perform during unprotected left turns and navigating through a roundabout?
-The FSD system performed well during unprotected left turns, slowly creeping forward to ensure oncoming traffic had cleared before proceeding. In a roundabout, it maintained a good speed and navigated smoothly, although it did not signal when leaving, which the narrator felt was appropriate to avoid confusing other drivers.
Outlines
๐ Self-Driving Progress in Berkeley Hills
The video script discusses the advancements in Full Self-Driving (FSD) capabilities, particularly in challenging environments like the Berkeley Hills. It highlights the system's evolution from being uncomfortable to use to now managing most of the driving tasks, albeit with occasional odd behavior. The narrator emphasizes the importance of unbiased news sources like Ground News for accurate information and showcases the car's ability to handle speed bumps, recognize stop signs, and yield to pedestrians. The autopilot visualizations are noted to have improved significantly, providing a more accurate representation of nearby vehicles.
๐ High Fidelity Visualizations and Defensive Driving
This paragraph focuses on the improvements in autopilot visualizations, suggesting that future updates may rely on more accurate models like those used in High Fidelity park assist. The video shows the car's defensive driving behaviors, such as caution around pedestrians and anticipating other vehicles' actions. The car's ability to handle narrow roads and oncoming traffic is demonstrated, along with its smooth navigation through tight spots. However, the script also notes some awkward behaviors, like stopping too early for a stop sign and issues with pulling away from the curb.
๐ Navigating Tricky Intersections and Speed Limit Confusion
The script details the car's navigation through an unprotected left-hand turn and its overly cautious behavior around pedestrians. It also discusses a situation where the car misinterprets a faded speed limit sign, causing it to drive at an uncomfortably slow speed. The car's difficulty in recognizing road closures and construction zones is highlighted, noting the need for further improvements in these areas. The car's decision-making process in various traffic scenarios, including yielding to oncoming cars and navigating roundabouts, is also covered.
๐ฆ Challenges with Road Closures and Construction Zones
This section of the script addresses the car's struggle with road closures and construction zones, leading to disengagements. The car's path planning and its reaction to road signs, including a 5 mph sign that caused a significant slowdown, are discussed. The car's handling of a private driveway and its inability to reverse on its own are also mentioned. The narrator expresses hope for improvements in future versions, particularly regarding construction zones and road closures.
๐ Navigating Construction and Road Obstacles
The script describes the car's approach to navigating through road closures and construction areas, noting that it sometimes struggles to find the correct path. The car's interaction with a construction worker holding a stop sign is highlighted, along with its ability to handle a roundabout and narrow roads. The car's decision to pull into a red zone instead of a proper parking spot is criticized, and the need for the car to learn to avoid such areas is emphasized.
๐ Smooth Driving and Successful Navigation
The final paragraph summarizes the drive, noting that it was the most comfortable and least stressful one through the challenging area. The car's full self-driving capabilities have improved to the point where it handles most situations competently, with only a few interventions and disengagements. The narrator expresses optimism about the progress of FSD and thanks the viewers for watching.
Mindmap
Keywords
๐กFSD
๐กDisengagements
๐กOccupancy Network
๐กHigh Fidelity Park Assist
๐กDefensive Driving
๐กPath Planner
๐กSpeed Limit
๐กRed Zone
๐กRoundabout
๐กConstruction Zones
๐กAutopilot Visualizations
Highlights
Stress Testing Tesla FSD V12 in challenging narrow streets of the Berkeley Hills.
FSD has evolved from uncomfortable usage to managing most situations with occasional stress.
Significant progress in FSD's ability to navigate complex roads compared to a few months ago.
Ground News is introduced as a tool for unbiased and comprehensive news coverage.
Autopilot visualizations have improved, rendering cars more steadily and accurately.
FSD's handling of speed bumps and stop signs has become smoother and more human-like.
The removal of cone visualizations and introduction of occupancy network visualizations.
FSD's cautious behavior around pedestrians and its ability to anticipate actions.
FSD's autonomous decision-making in navigating tricky road and traffic situations.
The car's adherence to road rules, such as yielding to uphill traffic, even without explicit programming.
FSD's occasional failure to use turn signals when it's the only vehicle on the road.
Challenges faced with road closures and construction zones, requiring human intervention.
FSD's difficulty in navigating a private driveway and resulting in a disengagement.
Smooth handling of steep and narrow roads with minimal steering wheel movement.
FSD's struggle with road closures and the need for improvements in upcoming versions.
The car's incorrect adherence to a 5 mph speed limit sign leading to slow and problematic driving.
FSD's overall improvement in comfort and competency despite remaining challenges.
The drive concluded with two interventions and two disengagements, highlighting areas for future updates.