Welcome to Tesla Motors Club
Discuss Tesla's Model S, Model 3, Model X, Model Y, Cybertruck, Roadster and More.
Register

The Mapping Singularity is Near: Medium Definition Maps

This site may earn commission on affiliate links.

diplomat33

Average guy who loves autonomous vehicles
Aug 3, 2017
12,700
18,660
USA
I thought this was interesting. The article discusses "medium definition" maps that would be in between standard definition maps and HD maps. The idea is to create maps that are more scalable like standard maps but still provide rich data similar to HD maps to be useful for autonomous driving. So, MD Maps could be the best of both worlds.

Here is the article: The Mapping Singularity Is Near

The article also a nice comparison of SD, MD and HD maps:

Standard Definition Maps

1*cR5194R6RJJJdMEk6nyiiw.png


Medium Definition Maps

1*GOAtRGmSyIE7mK1hNTry_g.png


High Definition Maps

1*WSj2xE7sj3b6Ea6uUNw6fw.png
 
Tesla mention:
Traditional HD map building no longer should be an impediment to greatly expanding L2+ hands-off operating domains for increasingly discerning car buyers. One of the obvious responses to this might be: Isn’t Tesla already doing this without maps at all? The answer is — not quite. While they have spoken publicly about their aversion toward “HD maps,” Teslas today do use higher definition data than found in a conventional SD map (e.g., more information on things like lane counts, turning options, traffic control). While many find their approach to “Full–Self-Driving” problematic (including the term itself), this leveraging of enhanced map information is useful to understand what is (and isn’t) possible with ~MD today.
 
  • Like
Reactions: diplomat33
Tesla mention: Traditional HD map building no longer should be an impediment to greatly expanding L2+ hands-off operating domains for increasingly discerning car buyers. One of the obvious responses to this might be: Isn’t Tesla already doing this without maps at all? The answer is — not quite. While they have spoken publicly about their aversion toward “HD maps,” Teslas today do use higher definition data than found in a conventional SD map (e.g., more information on things like lane counts, turning options, traffic control). While many find their approach to “Full–Self-Driving” problematic (including the term itself), this leveraging of enhanced map information is useful to understand what is (and isn’t) possible with ~MD today.

I think MD maps might be a good solution for Tesla since it would allow them to do very scalable maps (the big reason they hate HD maps) while also providing more data in their maps that could help FSD. In fact, I feel like Tesla might be doing their own version of MD maps since we know they are using maps with more data than basic standard maps but not at the level of HD maps.
 
Based on what verygreen has shown, their maps are alot more defined than what the author classifies as MD and more closer to HD than to MD.

I am not really sure about that. I don't doubt that Tesla's maps have a lot of features but if they were really HD maps, they would be more accurate and FSD beta would not make mistakes like driving in the wrong lane.
 
  • Like
Reactions: powertoold
I am not really sure about that. I don't doubt that Tesla's maps have a lot of features but if they were really HD maps, they would be more accurate and FSD beta would not make mistakes like driving in the wrong lane.

HD map won't stop a car from driving into the wrong lane. A safety driver does that the second the car even hints it gonna cross the line. Take a look at the disengagement description from CA reports to see what i mean. The reason you see Teslas cruising in the wrong lane all the time is because the planning system isn't good and because the drivers let it happen. The idea that HD map will instantly fix a-lot of Tesla's driving policy issues is simply not accurate. Not saying if they used accurate 10cm HD map that they won't be better. But that they are making simple mistakes that is the result of a poor planning system.
 
  • Helpful
Reactions: diplomat33