Smart Carts

What you are looking at is a low-speed
electric vehicle frame with 3D printed body panels.Hopefully we’ll release a
small fleet of these for on-campus travel so if you need to go from North
Campus to Central Campus or just between buildings far away or maybe
there is some VIPs that we want to impress so we dial up an automated golf cart to come pick them up. Being able to 3D print cars allows us to experiment with the shape of the vehicle.
How do we make the vehicle inviting? How do we, sort of, make the vehicle tell the story
that “Hey, I’m an autonomous vehicle come and ride on me!”. I am putting the
steering wheel on because it’s not self-driving yet so in order to move it back
to the lab we have two shamefully put the steering wheel back on.This is our first autonomous low-speed electric vehicle. We envision in two
or three years that we might have a fleet of fifty or a hundred of these
guys zipping around campus and just basically serving as a force multiplier
for the bus system that we have on campus. Wait… guys, is there a service
elevator that’s big enough for this? (grumbling) So a lot of people talk about
intelligent transportation systems that are a bit like using Uber where you pull out your
phone and an autonomous car comes and picks you up. How does an autonomous car figure out who
on the side of the road called it? How does it identify a safe place to pull
over? The goal of this project is really to tackle the transportation system
challenges in order to turn it into a viable economically and socially
acceptable transportation system. We’ve got a team of about sixteen people and it’s
really important to locate the team physically close to where you’re testing.
We started working at MCity yesterday morning this gives us a really
controlled test environment that is huge for rapid development. One of the
disadvantages is that there isn’t any actual structure at MCity like in
terms of contained buildings with running water and so we are in a
shipping container with a porta-potty but that’s what it takes to get a
cool test facility like this then so be it. Alright are you ready? I’m going to draw
a route for the vehicle to take, hand-free. We spent basically the first month just trying to
get the vehicle to turn the steering wheels and apply the throttle and the
brake when we wanted it to. I think the hardest technical challenges for us to solve is the localization problem. Being able to keep track of where the vehicle
is with respect to the roadway and which lane we’re in and all those sorts of things.
We’ve got a GPS system and that’s aided by encoders on the wheels and a fiber
optic gyro in the back seat we also have a laser based localization
system that is building a 3D map of the vehicle around us and then the third
method that we’re looking at is using radio beacon transponders so later today
we’re going to be populating about 50 beacons all-around MCity. It’s not
practical to instrument the whole world with hundreds of beacons but what we could do
is on a scale of like a campus-wide navigation system it is practical on
that scale. For a transportation on demand system
like this to be really successful it’s going to require a lot of people to work
together. If we’re successful though I think there’s a huge upside in terms of
social and economic benefits. Jinsang Kim told me he had some kind of super cool
liquid so I’m out here to ask him what makes that liquid so cool? When Professor Jinsang Kim and his team first made the new molecule, they found that it can be
either a liquid or …