I have been thinking a lot about decentralization and how technology can allow things to be managed by a network rather than a central authority. I keep going back to an idea that I heard somewhere (I can’t remember where but if I do I’ll post the link) about a decentralized taxi service of self driving cars. The idea is in a big city, the city government or Google or somebody provides the initial investment of self driving cars with software that allows the cars to manage themselves.
If you need a car you send a message from your phone, the nearest one comes to get you. When the car needs to charge its battery, it drives itself to a charging station. When it needs maintenance, it drives itself to a garage. Payments managed by Bitcoin’s blockchain, etc. Using this model, there doesn’t even need to be a taxi company.
This sounds like a great project for Google and or Tesla, but not for us “regular people.” For true decentralized power, things needs to scale down as well as up. That got me thinking about how to scale down this idea to a starting point that is achievable on a small budget and without legal constraints. Than I thought, quadcopter drones are getting cheap!
What if you start with a quadcopter that could do two things: 1. Anything useful, and 2. Charge themselves when they need a charge. I figure the useful thing should be something easy like ‘hover around while lighting up and looking pretty at night.’ I thought the cell phone chargers that you can just set your phone on and they charge wirelessly through induction might work great for this. Could a drone could just land on a pad and charge?
The answer is yes, and people are working on it. The problem is that even micro-quadcopters need quite a bit more juice than a cell phone so the commercial chargers won’t hack it. These guys built one that works, but its a bit larger and more technical than hobbiests could manage. Ok, so no wireless charging. What about a modular battery that can be replaced by a small robotic arm like the ones used in car manufacturing? Oh magic internet, what can you show me? Someone made this already out of legos!? Why thank you amazing future world we live in!
So now we have a starting point sufficiently scaled down. The question is, how do you scale up from the back yard to world changing? The following is how I envision that process roughly step by step. There are still plenty of gaps and details that will need to be filled in.
Most of these steps represent increasingly difficult software and hardware challenges. I am not a experienced programmer or engineer, so I honestly do not even know how difficult some of them would be. Some of these challenges have been solved already in other projects, many have not. If this project becomes a movement, as more and more people work on it the collective intelligence of the humans collaborating on the project will grow exponentially. Every step is possible. The more people work on it, the faster we can get to step 20.
Ok back to the backyard:
1. Start with a small, inexpensive autonomous quadcopter, maybe like this, and a robot who’s sole purpose is to charge batteries for its drone friends and swap them when they land.
2. Program your drone to know when its low on juice, and land before it dies and crashes.
3. Add the code to seek out its robot battery charger/ changer friend, and land on/in/next to it.
4. Redesign the battery pack to be 3d printable, and swap out easy. This system will be replaces later when induction charging works better and is cheaper.
5. Put pretty lights on it. Now you have a flying light up friend and his battery helper friend that you can pretty much forget about (until it gets buggy and crashes).
6. Get more of them and add code allowing them to talk to each other (swarm).
7. Get a solar panel to attach to the batter charger. (This will add to the autonomy).
8. Open source everything so other people can add more abilities besides “fly around and look pretty.” “Find litter, bring to trashcan” might be an early ability. Open source will also allow people to start their own swarms.
9. Program them to identify when they need maintenance, signal you before they break.
10. Program them to send you a text with their location if they crash.
11. Program them to send each other msgs with their location if they crash, and have one or of them airlift the malfunctioning one to you.
12. Add code so that they when they need repair they send a tweet. People in your area can volunteer to repair them. The drone will fly (or be airlifted) to the nearest volunteer. Other enthusiasts will put up their own charging stations/ battery-swap robots. Soon they won’t need you anymore.
13. As people build their own swarms, neighboring swarms combine and their networks grows. The bigger it gets the faster and more powerful it gets (as is the case with decentralized computing which is only one reason why we should build everything this way.) At this point they start lighting up the night sky all pretty all around the city and beyond. When the government comes knocking you explain that these are autonomous robots, and if they don’t like what they are doing they will need to take it up with them. If they ask which robot is in charge, try not to laugh directly into their face. Remind them that these are helpful and friendly robots, and all they are doing is looking pretty and picking up litter.
14. Lets get really ambitious. We add code that programs the swarm to evaluate the intentions of other drones. If they encounter a drone that is spying or otherwise doing something that has to do with controlling people, they wirelessly hack the drone, download their own software onto it, thus converting the brainwashed evil spy drones from the dark side to the light side, and adding to their numbers. This is actually not as far off as you might think. Here is a guy explaining his open-source software that is about half way there!
15. When people suggest their own software updates, if it is approved by the community (of people) it becomes part of the software that they all run on. New software gets downloaded onto one new drone, which gets sent out to join the others. When it arrives the software update jumps from drone to drone via wifi or bluetooth, like a helpful virus.
16. People add more capabilities like a delivery service. Now if you need something under x ounces delivered, send a message to the swarm. Same as if you needed a ride with the self driving cars mentioned above. Order some beer!
17. The community adds other capabilities. If you need arial footage of traffic, or flooding a disaster area? They can now do that for you. Maybe someone figures out how they can fly along powerlines and charge by induction while traveling. Now these things can be helpful in a huge number of ways.
The maintenance and charging may be automated, but it costs (digital) money. If you want something delivered or video taken, they will charge you bitcoin to an account. We have developed software that keeps track of the maintenance costs of the swarm. They charge only as much as needed to keep going. Services that benefit an individual (like a delivery) cost money. Services that benefit society (like the pretty lights and the litter pick up) are free. Now people that fix the drones in the growing swarm can be compensated. Expansion takes place when people like you and me build more and add their own, which is also compensated. This is a non-profit situation. Like many people, drones do not require a profit motive. They just need purpose, and a little TLC.
19. As they get smarter, they will become more capable on every level. This means they need to have good values, which means that we need to have good values. Everything about this system is about being helpful rather than harmful. “Do no harm” should be hard-coded at every level. I believe the open source community is up to the task.
20. As creator of the first swarm network, you have no special privileges. You have no more or less access to any of the code, or any of the drones themselves. But there are thankful people who send drones to deliver you gifts of fine cigars, artwork, and home made candies from time to time. Perhaps you accept Dogecoin donations (I do).
Why we should do this (besides the obvious fact that it would be so cool!):
We are leaving the age of the gun and entering the age of the autonomous drone. Governments and corporations are building armies that do not require that they win “the hearts and minds” of people.
In its maturity, this kind of network of robots could be used to protect people. In its infancy, it can change the conversation about robots and their roles in society.
Jobs are being automated quickly.
This project could accelerate the automation of jobs, but it would do so on our own terms. Utility and value will be maximized while cost will be reduced to near zero. Under the current system automation leaves people dangerously expendable and vulnerable. Under a new system, automation could leave us with richer lives.
We need to get ready for the future. I suggest that this is one way we can prepare. We create our own “for the people, by the people” network of robots that reflect the values of the world we want to live in. Decentralized, open, connected, intelligent, and collaborative.
Lets start building!