Autonomi: The Anti-Uber

How blockchain enables a truly autonomous, self-governed car-sharing platform.

Written by Charles Wong and Aravind SK.

Charles is a self-driving car engineer and sustainable energy advocate. He loves robotics, AI and wants to build a space-faring future for humanity.

Aravind is a self-driving car engineer and an AI researcher. He believes in augmenting the human experience through the convergence of neuroscience and AI.

Together we’re building a DIY self-driving car, but that’s a story for another time.

This is part 1 of a series about Autonomi, the first truly autonomous ride-sharing platform with minimal human governance.

Courtesy of Nvidia

Humans are an incredible species. We invented the microprocessor, a marvel of modern technology. We discovered science, music and art, hallmarks of civilisation today. We sent a man to the moon, and soon, we might be living on Mars.

Yet for some unfathomable reason, we collectively decided the best way to transport a 70kg bag of flesh and bone (us) was to build an almost 2-ton steel behemoth that runs on exploding dinosaurs.

The combined carbon footprint of private vehicles around the world is unsustainable. Our mobility situation has to change and it has to change fast.

Happily however, companies like Uber have a solution.

Uber’s Vision

In 2009, Uber was founded with the goal of connecting private car owners with eager commuters. The idea was that commuters would find in Uber a cheaper, more accessible way to get around than taxis, and private car owners could make use of the extra space in their cars.

Benefits included added income for car owners and cheaper travel for commuters. This would result in less cars on the road, lower congestion and a possible path to a carbon-light future.

The reality hasn’t been so peachy however. Transportation Network Companies or TNC’s have caused a net increase of 600 million vehicle miles traveled in New York City — a 3 to 4 % jump in citywide traffic. San Franciso transportation agencies have been pressing the California Public Utilies Commission to reveal Uber and Lyft trip data, with concerns that the increased vehicles traffic caused by the companies is having a significant environmental impact.

Despite this, ride-sharing services are flourishing. Uber is worth an estimated $69 billion, and Lyft is worth $7.5 billion at time of writing. That’s great and all, but the services’ popularity and reach come at a cost. We believe this model simply cannot sustain itself over the long term, and is built with profit first and a better world second. Reasons include:

  1. Ride-sharing services are adding to existing congestion — drivers often purchase or rent new cars for their work
  2. Ride-sharing services do not actively reduce private car ownership
  3. Private car traffic has risen since Uber’s introduction
  4. Ride-sharing services take a significant cut of the fare (20% for Uber) and leave drivers to bear the cost of vehicle maintenance

The Uber model is far from perfect. We think there’s a better way.


Cars Run the Show

Here’s our proposal. Imagine an autonomous car in New York. It earns money by being a private taxi, and services its customers through a ride-hailing app. When it runs low on power, it searches for the nearest available charging station to recharge. If it needs a service, it searches for the nearest mechanic for a tune-up. The car’s onboard AI decides how best to keep itself in service through regular maintenance and upkeep, all without human intervention. With me so far? Good.

Now things get interesting. Let’s say the car has made enough money. It can decide to buy another autonomous car, expanding its fleet size to two. The first car tells the second car to also behave like a taxi, and they both continue to service their customers. Over time, the fleet grows until it’s able to meet all demand in its locale at all times of the day. The fleet can activate only the vehicles it needs, reducing congestion in crowded city centers.

Unlike human drivers, the cars don’t need to rest, and are capable of running 24/7/365. This means the system can be more effective than any current ride-sharing platform while using less cars.

The cars can charge customers lower fees as they don’t need to pay their drivers, nor give their human administrators a significant cut.

They operate only to serve their customers and keep themselves functional — nothing more.

Cheap, accessible, unbiased private travel available to the greater public. We think a system like this is incredibly exciting. It may seem a bit of a pipe dream right now, but the technology is right at our fingertips, and can be made reality sooner than we think.

System Requirements

The most glaring problem with this system is the vehicles themselves. Level 5 autonomous vehicles are some ways away, and it’s not clear when they’ll be widely available. That said, this is a problem that’s being solved right as you read this. The biggest names in the space like Waymo, Tesla and even Uber themselves are working to make self-driving cars the de-facto standard on the road.

At World.ly, we think the next biggest challenge isn’t with the self-driving technology. It’s with payments to and from the cars themselves. Established payment systems like PayPal and VISA exist, but are less than optimal for a network of cars plying the streets. Passenger fares, congestion charges, road tax, repair costs, accident settlements and debt payments are just some of the hundreds upon thousands of transactions a car may have to deal with everyday. The long processing times and high fees of payment gateways simply can’t cut it. These expenses are always passed on to the customer, and the only parties that benefit are the payment companies themselves.

Blockchain to the Rescue

What we need is an fast, free and secure way of processing those transactions. That’s where blockchain comes in. At its core, the blockchain provides a decentralised, distributed ledger capable of processing transactions across the planet at a fraction of the cost. More mainstream currencies like Bitcoin and Ethereum would work reasonably well for this, but lack the instant speeds such a high volume system demands. Their volatile nature is also less than ideal as stable stores of value, but that may change as the cryptocurrency market matures.

After extensive research into the available blockchain platforms, IOTA stood out as a prime candidate for this application. In fact, we wish we’d thought of IOTA ourselves. It touts itself as the backbone of IoT for the autonomous Machine Economy. It offers instant payments of exact amounts between entities completely free. They achieve this by their proprietary protocol Tangle, an ingenious concept that improves the more entities join it (more on that in part 2). In the context of payments between customers, cars, mechanics and charging stations, IOTA sounds like the perfect match.

With cumbersome traditional payment gateways out of the way, the cars themselves are free to behave as independent machine entities, with greater financial fluidity without having to think about transaction fees. This allows them dynamically meet the demands of a city at prices lower than any traditional private transport system could offer, with far fewer vehicles on the road.

What Else?

One of the most prevalent problems in ride-sharing is the uncertainty in availability and pricing. Rides aren’t always there when you need them, and if they are, you get overcharged for it.

Most people follow a routine travel schedule. They come and go at the same time everyday, and their cars sit idle in the meantime. What if a few parties with complementary schedules could pool to adopt a car full-time?

Alex works in the city and travels to and from his home in the suburbs in the mornings and evenings. Stacy is a freelancer, and only needs her car in the afternoons when she’s meeting a client. Daniel loves the nightlife, and joins his friends for drinks every other day.

It’s plain to see how the one car built to service a select few parties at different times of the day could maximise usage and transport efficiency. All parties pay significantly less, and the car is being used more of the time.

We envision an ideal future where your car learns your travel habits, predicts where you’ll need to be and simply shows up at your doorstep.

This would not a replacement for ride-sharing or private car ownership. It occupies a unique space in between both, for people who may not need the car for the whole day, but still want the stability and certainty of private car ownership at a much lower price point.

A Better World

Elon Musk thinks firing rockets around the planet to get us from Paris to New York in under half an hour is the future of continental travel. We think having a truly autonomous, decentralised seamless transport platform is the answer to our urban mobility problems.

The Uber model is wasteful, and driven by the need for ever-higher valuations and global market capture. We want to provide the world with a democratised means of transport that serves humans in the most efficient way, at the lowest cost and with the least environmental impact. We think the marriage of artificial intelligence and blockchain technologies is our fastest ticket to that ideal. We’re going to build it. We’re calling it Autonomi.

Stay tuned for part 2, where we break down the underlying technologies behind Autonomi, and what we think is needed for the platform to succeed!

Like what you read? Share your thoughts below!

Think this idea is worth spreading? Share it with your friends and co-workers!

Find us on Github! https://github.com/World-ly/Autonomi