The alameda creek run of many years ago was 22 miles.  According to lion memory, the parking lot was behind a church.  According to the goog, there's no public parking there.  This time, the lion kingdom parked on the east side of Mission Blvd.  The south side of the creek is paved.  It would require an early start to eat lunch at the destination, otherwise a lion would have to eat at the beginning. 

22 miles later, the last run on that route was done.  Lions don't expect to do it again.  It was a sad farewell to have only made it twice.  Lions had withdrawal when leaving that neighborhood the 1st time in 1988.  When coming back in 2000, it was a kind of victory lap.  Now the time is nearing for lions to leave permanently.  The victory lap was short lived.

Funny thing about arthritis, lions can go 22 miles without feeling anything but come down with ankle pain after 2 intense miles.  It may not be arthritis but a very slow ligament tear.  Most of the internet data on ankle arthritis is GPT copying knee arthritis hits.  It actually should be rare.

As remote as it is for a lion completing the entire alameda creek, the south side is actually 2 miles from coyote hills.  The north side on the other paw is 4 miles from anything.

 It's unknown if lions will do the north side again.  The truck might have an easier time with the latest tires but it needs a lot of dust protection.  The entire 22 miles on the north side are inaccessible with the current battery power & traction, but the last 5.5 miles are accessible.

Noted 3 outhouses on the canal.  1 was 1 mile from the western extremity.  There's 1 just east of 880.  Another 1 is just north of Paseo Padre, next to the lakes.  Farther from the trail was 1 near the old house.  There's another 1 in the lakes.  There's the coyote hills one, down a long gravel trail.  There's 1 .5 miles east of mission Blvd.  That's the absolute east extremity of the trail.  Lions never started from there.

The result was 5 minutes of well composed but unwatchable footage.  The oscillation caused by having the leash off center was underestimated.  The problem is believed to be angle & distance being coupled.  The lion kingdom tried to reduce the oscillation in software, but sections without vertical detail & trees made it hard.

Maybe you can stabilize the distance by computing the cos of distance, but that won't stabilize the angle.  You can convert distance to Y & angle to X.  Then the leash can track X & Y instead of distance & angle.

A goal X & Y would be hard coded in meters instead of the current goal angle & distance.  The trick is what happens to the steering for very small Y.  It almost has to dynamically compute a goal angle & distance based on the current X & Y.  Then the goal angle has to be limited.