Saturday, January 7, 2017
Autonomous bike path snowplows
Attention Mayor Peduto, PennDOT, Uber, Google Robotics, Bike-Pgh, and others with a stake in winter cycling in the city:
All this talk about autonomous cars and taxis and so forth is missing an important boat. Cycling after a snowfall in Pittsburgh is darn near impossible because the city, or PennDOT, or whomever, does not ever, it seems, clear the bike paths. This needs to happen hours after a snow, not days or weeks or not at all. I myself am the proud owner of a busted shovel, accomplished by chopping ice off of the ramp from Second Avenue to the Hot Metal Bridge in February 2010. Nobody even tried to clear that ramp between the Feb 6 storm and the Feb 20 shovel brigade. To this day, if it gets done at all, it is long after a snowfall.
The solution is to have a set of brush trucks traverse the bike paths, in the middle of the night if necessary, anytime it snows. If humans are not available to make this happen, it should be done by automated driving, like the Uber cars. Imagine, snow starts to fall at midnight, but at 3 a.m., the automated plows come to life, and begin their routes. By daybreak, the paths have had at least one good pass. If it's stopped snowing, they return to their sheds (or a fueling station), but if not, they repeat their routes.
This will require some figuring out. Unlike streets, bike paths do not follow the same rules of pavement markings or lane widths. Still, it's a solvable problem. For example, if cutting through deeper snow, it may not be possible to discern where the path is at all, relying exclusively on GPS coordinates.
But high-tech aside, even if a human was available to do the driving, it's still going to take a fleet of these to handle clearing our many miles of bike paths, bridges, ramps, and other infrastructure. Fuel, routine service, and longer-term repairs, all seem like the sort of thing that has already been figured out for other applications, say PennDOT keeping the grass trimmed along highways, or street sweeping equipment owned or contracted by any of the above. This shouldn't be that difficult to plan and execute, provided the money is available. (Separate discussion.)
Can we, sometime this winter, develop a plan for getting this done? Give each machine a range of, say, eight miles to clear, so four miles out and back. Some that come to mind:
* Eliza Furnace Trail and Smithfield Street Bridge. (route)
* Bottom of Boundary Street to Hot Metal and its ramps to the end of South 18th. (route)
* 6th, 7th, 9th Street sidewalks and bike lane, if present, plus ramps to river level (route)(only about 4 miles, but many street crossings, plus heavy usage)
* Fort Pitt and Fort Duquesne Bridges (incomplete route)
* Penn Ave bike lane, starting at Point State Park, looping at north end of 16th St Bridge, and Spring Way out to 31st Street (route)
* North Shore Trail, from Alcosan to Millvale (route)
* South Side Trail, from Duquesne Incline to South Side Works (route)(SSide works is about 4 miles of the 9 shown)
* Strip District Trail, and along the city side of the river from Convention Center to Point State Park (route)
That's eight paths, just off the top of my head, and I'm sure there are plenty of others. I didn't try to optimize this, just sketched a few ideas. We should come up with a more scientific approach to this, complete with actual humans with appropriate equipment, to perform a full sweep after a couple inch snowfall. Get some data, figure out what it's really going to cost to acquire and run this service. Put it out there when we get a stronger storm, say 6" or more, see whether or where it gets bolluxed up.
If we care enough to mow the grass in our parks in the summertime, we should care enough to sweep the bike paths in the wintertime.
This idea gets bandied about every winter, yet nothing happens. How about this winter, we finally do something about it?