Courier Network (For real)

[kiba]

Hmm..I am thinking of starting a robotic courier network...for real!

I think the first step is to start small, yes?

Delivery of USB drives is ideal! The first route between two bitcoiner that is the shortest distance will have to be determined.

That mean:

1. We can start with a small robot prototype..and can be used to conduct the first courier transaction!

2. Start in an area where bitcoiners are unusually concentrated such as New York and Washington, D.C. area!

3. Start with delivery of a cheap but still valuable good.

[benjamindees]

Use a blimp.

[Anonymous]

Use a blimp.

 a second hand predator drone.

or
http://www.suasnews.com/2011/03/4720/hexacopter-that-detects-motion-and-breathing/

[Gryph]

A small autonomous robot capable of carrying a useful package (say ~1kg) and having sufficient sensors and machine vision capabilities to guide itself from point A to point B and avoid obstacles, vandals, unforeseen occurrences (bird hitting it or a kid running in front of it), and ward off the occasional angry house pet...

This is going to cost a LOT.

Since they have a limited lifetime and require fuel and maintenance would the total costs ultimately be cheaper than a minimum wage dude with a truck?

[Nefario]

A small autonomous robot capable of carrying a useful package (say ~1kg) and having sufficient sensors and machine vision capabilities to guide itself from point A to point B and avoid obstacles, vandals, unforeseen occurrences (bird hitting it or a kid running in front of it), and ward off the occasional angry house pet...

This is going to cost a LOT.

Since they have a limited lifetime and require fuel and maintenance would the total costs ultimately be cheaper than a minimum wage dude with a truck?

Yeah, I think dude wth a bycicle/truck/car/running shoes would be much more realistic starting off.

Apart from the reasons above, not many people in NewYork or Washington have landing strips.

Once bitcoin becomes self aware we don't want to give it a head start with drones now do we.

Having a look at the bitcoin map the biggest concentrations of bitcoin users is on the upper east, Boston, Mass., NY etc. And in L.A. and San Fran.

There are also quite a few around Florida, a lot in Dallas, Collarado for some reason, Utah, and near the coast in Washington state.

I am sure an informal local delivery service can be set up by those users, that then becomes a node, and every once in a while nodes can send large groups of packages to each other.

[kiba]

A small autonomous robot capable of carrying a useful package (say ~1kg) and having sufficient sensors and machine vision capabilities to guide itself from point A to point B and avoid obstacles, vandals, unforeseen occurrences (bird hitting it or a kid running in front of it), and ward off the occasional angry house pet...

This is going to cost a LOT.

Since they have a limited lifetime and require fuel and maintenance would the total costs ultimately be cheaper than a minimum wage dude with a truck?

Who say the robot have to be...autonomous? It could start with a simple prototype.

[Gryph]

Who say the robot have to be...autonomous? It could start with a simple prototype.

Wait.... so we have to pay for a skilled-above-minimum-wage dude to drive it, AND pay for the robot itself?

A dude with a truck sounds cheaper.

Unless you're actually proposing to build a truck and just giving it a fancy name of "robot".... that *might* be cheaper.... but ultimately thats like FedEx making their own vehicles...

... i'm not saying its impossible, but complete "vertical integration" of a business to *make* the stuff they need to do their business I've only seen work when the stuff actually doesn't exist yet... (Like NASA...and that's not even a profit making business).

[kiba]

Wait.... so we have to pay for a skilled-above-minimum-wage dude to drive it, AND pay for the robot itself?

We are talking about robots that:

1. Drive short distance between two locations.

2. Pedestrian speed.

3. Poor performance characteristics.

That's what I have in mind. I don't have in mind a Serious Business with godly computer programmers cranking out algorithms.

Rather, it's composed of a network of robotic courier enthusiasts who sometime got paid to deliver a few goods there and there and the network will start extremely small. Eventually, we'll start learning how to create economical designs and the software will be of better quality.

(That being said, high quality sensors are getting cheaper and we're becoming more knowledgeable about machine vision)

[Nefario]

I can see these non-flying, walking speed robots becoming victims of children and bored teens that populate their route.

You might have a look at how courier robots are used in a hospital. They have been minorly succesful but have turned out a little expensive and a little unreliable.

And their unreliability has nothing to do with the robot itself, but with people in the hospital just stopping them out of curiosity(or kids).

Seriously do this with people before you start using machines.

[kiba]

Seriously do this with people before you start using machines.

Because they sucks so we should never use machines?

[kiba]

Anyway, I am more interested in the robotic than the courier part.

[Gryph]

Anyway, I am more interested in the robotic than the courier part.

I think then you should focus on probably with a robot can specifically do better than a human.

A courier service is one of those things that humans are naturally really good at. They have built-in sensors, obstacle and collision avoidance algorithms, their own propulsion system (takes up nearly half their body), energy consumption and distribution systems, AND they can deal with many unforeseen circumstances including angry house pets. Best of all, their initial base cost of around $110k-$150k is not paid by you, but by a parent human or even partially subsidized by the government. You just have to pay a rental fee that works out to approximately minimum wage.


Robots on the other hand are better for really high speed, excessively and rigidly repetitive, or extremely dangerous tasks. I think that's what you should go for.

Maybe design an ASIC that mines bitcoins even better than GPUs?

[kiba]

A courier service is one of those things that humans are naturally really good at. They have built-in sensors, obstacle and collision avoidance algorithms, their own propulsion system (takes up nearly half their body), energy consumption and distribution systems, AND they can deal with many unforeseen circumstances including angry house pets. Best of all, their initial base cost of around $110k-$150k is not paid by you, but by a parent human or even partially subsidized by the government. You just have to pay a rental fee that works out to approximately minimum wage.

Google cars are changing all that.

Maybe design an ASIC that mines bitcoins even better than GPUs?

BOO-ring. I am uninterested in mining.

[RodeoX]

I think we will one day see something like this. There are huge advances in robotics underway. This year I'm going to this event to check it out for myself.
http://www.robocup2011.org/en/

Or you could use one of these.
http://www.youtube.com/watch?v=ELr3U4l5HkQ

[Gryph]

Google cars are changing all that.

Wait.... So you're competing with google cars?


Good luck!
 
(snicker)

[kiba]

Wait.... So you're competing with google cars?


Good luck!
 
(snicker)

What I am getting at is that advancement in technologies is making it practical.

I am irate about people shooting up baby step plans for getting a robotic courier network operational. You gottach learn how to crawl before you can run. Trying to make robots do things that it wasn't good at is the whole point.

[Anonymous]

Cyborgs are theoretically robots.

 

[­­­Atlas_]

Decentralized delivery network?

Brought to you by Silk Road.

[Gryph]

What I am getting at is that advancement in technologies is making it practical.

I am irate about people shooting up baby step plans for getting a robotic courier network operational. You gottach learn how to crawl before you can run. Trying to make robots do things that it wasn't good at is the whole point.

Baby steps? what baby steps? you want something that can compete with human beings for an objective that is complex enough even for a human, (kids get lost all the time after all).

that's like saying you want to run a marathon on baby steps. This is why we're shooting you up.

If you want baby steps, instead define baby distances for the objectives.

Instead of delivering stuff in a real city. Make a much smaller baby objective of delivering an item around the corner of a hall way. Much smaller and more realistic. Then increase complexity as needed.

Your research plan would be MUCH more sensible and reasonable if a series of baby objectives were included rather than the final objective.

There is nothing wrong with baby steps, but no one unleashes their babies on the roads, they first attempt those baby steps in the safety of baby's home.

[kiba]

Your research plan would be MUCH more sensible and reasonable if a series of baby objectives were included rather than the final objective.

There is nothing wrong with baby steps, but no one unleashes their babies on the roads, they first attempt those baby steps in the safety of baby's home.

What we're arguing about here, anyway? We agree on the baby steps idea, it's just that I didn't make it sufficiently baby steps.