Show HN: Self-delivering autonomous bicycle | Hacker News Reader

Show HN: Self-delivering autonomous bicycle

2 pointsboxcardavin3 comments

I interviewed for the YC summer batch yesterday with my cofounder justincorbett for our company Weel, we did not get in. The idea is that you order an ebike on your phone, it delivers itself to you, you ride it like a regular electric bicycle, and when you are done it rides itself away autonomously.

The feedback we got during the interview and in the email last night was that the idea was interesting but the partners were not convinced that this would become something many people would want given how many other alternatives people had for last mile transportation. Basically, it's overteched for the problem.

Today we are regrouping and taking a fresh look at our own biases and views. I'd love any feedback from HN folks on the idea itself, the feedback we got from YC, and any criticisms or ideas you might have.

Landing demo with absurd music: https://youtu.be/hXDXjav3XD4

View on HN

Two cents...

+ What about last mile autonomous delivery?

+ You might want to have a recumbent bike rather than an ordinary, so people do not immediately see that no one is in there and you can transport stuff in it.

+ The bike shape is thin compared to four wheel options, is allowed to go in places cars aren't allowed and is easily recognized by road users.

+ Challenge will be the autonomous part just as with cars.

+ Pizza delivery... What's actually the cost to have a human doing that? Start in a region where this is really high.

Interesting, hadn't considered a different bicycle shape so that it looks like it isn't empty at a glance. Will look into this.

The delivery is totally possible, it doesn't really even take extra tech. The key insight we've found is that cities don't really care about helping make delivery cheaper, but they deeply believe in improving transportation. The city planners we've spoken with love this because they want to be free to design cities that aren't car centric, and to do that they need a massive shift from car usership.

Big tech benefits/savings over autonomous car computing reqs; it can stop on the side of the road if it gets confused, if it crashes no one dies, our speeds will always be low

Thanks for the feedback.

> "if it crashes no one dies"

That's a very optimistic assumption. A bike crashing into a pedestrian can definitely kill them if it knocks them over and they hit their head on something hard. Lots of elderly people even die of complications from broken hips. And a bike, even at low speed, can do a lot of damage to a small child.

A bike that swerves in front of a car can easily cause the car to lose control and crash into a person or another car, resulting in death or serious injury.

For these reasons, I'd expect that getting self-driving bikes to work safely (and to be allowed by regulators) would have most of the obstacles that self-driving cars face.

Note that in certain places (e.g., NYC), bicycles are subject to all the same traffic laws as cars. For example, they must stop at red lights, yield to pedestrians in crosswalks, respect one-way streets and not ride on sidewalks.[1] (It's not clear whether it would even be legal for an autonomous bike to drive in a bike lane, since it's certainly a motorized vehicle when driving autonomously.) Thus, an autonomous bicycle would have to have the navigation skills to be able to safely share streets with cars, trucks and buses in dense urban traffic. Can an autonomous bike carry around enough computing power and sensors to handle those kinds of situations?

[1] While most bicyclists in NYC ignore traffic laws, a self-driving bike that's not even capable of obeying such laws would certainly be frowned upon.

Reply on news.ycombinator.com