Uber Case Study — The State of Autonomous Transport
Uber ATG(Advanced Technologies Group) is dedicated towards advancing research and building technology to propel Uber in the Autonomous vehicle space. As I highlighted in my previous article, where I explored this space, ride hailing startups have a unique distribution vantage point when it comes to taking this technology mainstream.
Uber is a two sided marketplace of drivers(supply) and riders(demand). All of the different kind of rides that Uber offers is a way to provide more options to add supply and demand. For Uber the key KPIs include ‘average pick up time’ for riders and ‘average gross $ per hour’ for drivers. As Uber’s execution and network effect optimizes the utilization in the marketplace, the wages for the driver increase while the average cost to the rider keeps reducing. Uber does all this while collecting data for each ride. This trove of data is a moat which it leverages to move horizontally and vertically in this space. The former improves the data in the network while the latter will further optimize the supply. These two components will have an exponential effect on the network in the marketplace.
For the vertical move, data provides the direction for optimization (like improving average pick up time) while Artificial Intelligence and Autonomous technology are employed to achieve it.
National Highway Traffic Safety Administration (NHTSA) classified autonomous cars into five levels based on autonomous capabilities, which provides a framework for charting technological progress on a scale.
Level 0: No Automation: Driver controls all the functions of the car. No Automation. — These are most of the cars today without any automated functions.
Level 1: Function Specific Automation: One or more specific control systems are automated. For example, parking assist or lane assist. — The newer cars which came out recently have these functions.
Level 2: Combined Function Automation: At least two of the control systems are automated in tandem. For example, adaptive cruise control with lane centering. — This is where Uber and most of the other autonomous vehicle startups will start.
Level 3: Limited Self Driving Automation: All the control systems like steering, brakes and throttle are automated to work in tandem. The car monitors conditions to require relinquish control back to the driver. — Early versions of Cruise and other retro fit autonomous startups tried to approach this Level.
Level 4: Full Self Driving Automation: All control systems are fully automated requiring no human driver. — Final goal for most autonomous vehicle startups.
Uber is striving to go from a Level 2 to Level 4 with its data and technology prowess. We are exploring the autonomous landscape of Uber from the lens of their existing patents relating to autonomous technology. Using the terminologies we explained in our previous article we will try to classify the existing Uber patents.
Key Trends
1.Deep Vertical Expansion: Uber is growing its stack of patents in different categories to control the entire stack.
2. Acquisitions: Uber acquired deCarta and Microsoft’s mapping division to improve their mapping and navigation functionality. Uber also acquired Otto — a self-driving truck startup.
Key Observations:
I. Ride Hailing Centric Approach: Uber’s approach to autonomous vehicles is optimized for ride hailing. Patents concentrate on last mile delivery and provide autonomous solutions to problems in their existing ride hailing business.
II. Extensive testing combined with data: Some patents indicate Uber’s extensive testing combined with ride data giving them new insights for improving their autonomous technology.
Sources:
1. Data: Uber ATG, Uber Data Blog, Google Patents.
2. Landscape Detail: Autonomous Transport Landscape. (Previous article)
Disclaimer: Data for this study is sourced from freely available sources. Sole motivation for this study is to increase personal understanding of the autonomous car space. We aim to make this a living document, please share any thoughts and feedback and we will update this on a periodic basis. Feedback on any omissions, factual errors are also welcome.
Subscribe to our Acing AI newsletter, I promise not to spam and its FREE!
Thanks for reading! 😊 If you enjoyed it, test how many times can you hit 👏 in 5 seconds. It’s great cardio for your fingers AND will help other people see the story.
