Autonomous vehicles are no longer a side experiment in ride-hailing, as we explored in our previous article on Uber and Pony.ai. Lyft is now moving along the same curve, focusing on its role as a platform that connects riders with both human drivers and autonomous fleets.
Founded in 2012, Lyft has grown into a leading ride-sharing company, serving millions of riders while expanding into micromobility, premium transport, and platform-based mobility services, and its focus on autonomous partnerships represents the next step in that evolution.
Lyft’s major collaborations
Lyft’s autonomous ambitions are taking shape through a series of strategic partnerships that extend its global reach. In the United States, Lyft and Waymo plan to launch robotaxi services in Nashville in 2026, with Lyft handling marketplace operations and Waymo supplying fully driverless vehicles. In Europe, Lyft has partnered with Baidu to deploy Apollo Go autonomous vehicles across key markets, including Germany and the United Kingdom, marking one of its largest international moves into autonomous mobility.
At the same time, Lyft’s acquisition of Free Now is supporting pilot programs such as Level 4 autonomous taxi testing with the City of Hamburg, strengthening its role in public-private AV integration. Together, these efforts reflect Lyft’s transition from early experimentation to a platform designed to support mixed networks of human driven and autonomous vehicles at scale.
Lyft: Patenting Activity
From 2015 through 2019, Lyft invested in autonomous vehicle research, route optimization, and data‑driven improvements to its ride‑hailing platform. These efforts included refining core matching algorithms and exploring reinforcement learning methods to improve driver‑to‑rider matching efficiency. Lyft also partnered with autonomous‑vehicle developers such as Motional, conducting early public autonomous ride tests in Las Vegas and other pilot programs using Motional’s vehicles on the Lyft network.
In 2021, Lyft sold its Level 5 self‑driving division to Toyota’s Woven Planet Holdings for about $550 million, with payments structured over time. The deal included non‑exclusive commercial agreements allowing Woven Planet to use Lyft’s system and fleet data to support automated driving development.
After that sale and into the early 2020s, Lyft shifted toward partnership‑led technology strategies, and its patent activity declined accordingly. In 2024 and 2025, Lyft announced multiple new autonomous vehicle partnerships with Mobileye, May Mobility and Nexar to connect autonomous vehicles to the Lyft app and bring AV rides to Atlanta and beyond. Lyft’s partnership with Waymo also positioned its Flexdrive fleet management arm as a key operator for AV fleets without building the AV technology itself. In parallel, Lyft pursued AI integrations, such as a deal with Anthropic to improve customer service, showing broader tech engagement.
Lyft: Top Technology Areas
Lyft’s patent portfolio is anchored in information and communication technologies, with the highest concentration in G06Q and G01C. These areas reflect a strategic focus on data-driven orchestration of services, pricing, and platform operations, supported by precise positioning and navigation. Together, they highlight Lyft’s core strength in translating real-time data into efficient, scalable marketplace decisions.
Looking ahead, this foundation positions Lyft well for AI-driven optimization and greater platform automation. The portfolio’s depth in digital data processing (G06F), wireless communication (H04W), and perception technologies (G06V) supports future advances in autonomous and semi-autonomous systems, including vehicle coordination, sensing, and real-time decision support. In parallel, strengths in traffic control (G08G) and system-level management suggest optionality beyond ride-hailing, such as logistics, fleet management, and urban mobility services. Overall, Lyft’s IP strategy points to a platform built for intelligent automation, with the flexibility to extend into autonomy-enabled services and adjacent mobility markets as those technologies mature.
Designing routes for autonomous vehicles based on risk and exposure
As cars become increasingly self-driving, one major challenge is determining how safe and reliable they really are on different roads and in different situations. Simply counting how often the system switches off can be misleading, since many of these handovers are expected or would not have caused an accident anyway. That makes it difficult to assess how well the technology is truly performing or where it is safe to deploy self-driving cars.
Lyft’s patent describes systems and methods that process a ride request by collecting transportation details such as the pick-up and drop-off locations. The system generates two possible routes: one for a human-driven vehicle and another for an autonomous vehicle, with the autonomous route constrained by the fleet’s operating design domain. Unplanned disengagements are analyzed using human review or simulations to estimate likely outcomes, and these results are used to create safety and performance metrics for individual road segments.
These metrics are weighted based on how often each segment is traveled and the level of exposure. The autonomous route is then evaluated against the non-autonomous route to ensure that each road segment meets minimum safety thresholds, while also factoring in efficiency, distance, travel time, and passenger comfort. Once the optimal route is selected, an autonomous vehicle from the fleet is assigned to the request, supporting safer and more informed deployment of autonomous vehicles.
U.S. Patent No. 11,953,333, titled “Systems and methods for autonomous vehicle performance evaluation”, was filed on December 2, 2019, and was granted on April 9, 2024. The patent lists Kumar Hemachandra Chellapilla, Emilie Jeanne Anne Danna, David Tse-Zhou Lu, Sameer Qureshi, and Alexis Weill as inventors. Lyft was represented in the filing by James Soong of Sheppard Mullin Richter & Hampton.
Managing pickup and drop-off selection for autonomous ride requests
A key challenge in integrating autonomous vehicles into ride-hailing services is managing ride requests, pickups, and drop-offs in a system originally designed for human drivers. Riders are accustomed to giving instructions directly to a driver, and autonomous vehicles cannot interpret such instructions in real time.
Lyft’s patent describes a transportation system that receives ride requests from a user device, including pickup and drop-off locations, and identifies the geographic area associated with the request. It retrieves autonomous routes for the area and evaluates mapping, driving, and roadway conditions to determine whether autonomous vehicles are eligible to operate on the route. If eligible, the system presents autonomous vehicles as an option among other vehicle types and, when selected, matches the ride request to an available autonomous vehicle, sending instructions for the vehicle to travel to the pickup location.
Graphical user interfaces allow riders to select precise pickup and drop-off points, which the autonomous vehicle follows, and the system can offer alternative ride types if an autonomous vehicle is unavailable. This approach enables safe and efficient deployment of autonomous vehicles while preserving flexibility and user control in mixed autonomous and human-operated environments.
U.S. Patent No. 12,516,940, titled “Autonomous vehicle pickup and drop-off management”, was filed on June 6, 2023, and was granted on January 6, 2026. The inventors listed are Taggart Matthiesen, Sebastian Rolf Johan Brannstrom, and Jess Garms. The patent filing was handled by David Wu of Baker Botts.
Communicating vehicle intent to other vehicles and pedestrians
One of the main challenges of introducing autonomous vehicles into everyday traffic is that removing the human driver breaks many of the usual ways people communicate on the road. Without those familiar cues, pedestrians and other drivers may feel unsure about what the vehicle will do, particularly in complicated settings like intersections, crosswalks, and shared lanes.
Lyft introduces a transportation system in which sensors on autonomous vehicles detect nearby vehicles and road users and interpret how they are moving. A ride-matching system then coordinates these movements by exchanging messages between vehicles, allowing them to plan actions in relation to one another, such as waiting to let another vehicle proceed first.
The system can also communicate directly with pedestrians, cyclists, and other road users through notification devices on the autonomous vehicle, making its intentions clear. By replacing human-to-human communication with vehicle-based signals, this approach reduces uncertainty and supports safer, more efficient navigation in mixed traffic environments.
U.S. Patent No. 12,277,860, titled “Autonomous vehicle notification system”, was filed on November 30, 2022, and was granted on April 15, 2025. The inventors listed are Taggart Matthiesen, Jisi Guo, Sebastian Rolf Johan Brannstrom, and Jess Garms. The patent filing was handled by David Wu of Baker Botts.
Lyft: Top Law Firms
Lyft uses a small number of law firms and attorneys for most of its patent filings. Keller Preece PLLC handles the largest share, with partner Gregory Jolley closely tied to that work, showing a long-term and central role in Lyft’s patent efforts. Baker Botts L.L.P. is the second most active firm and supports a significant portion of Lyft’s patent filings. Darrow Mustafa PC handles fewer filings, but its work, mainly associated with attorney Christopher Darrow, shows an ongoing and consistent relationship.
Lyft also works with a broader set of firms for more limited or specialized needs. These include Sheppard Mullin Richter & Hampton LLP, Lee Sullivan Shea & Smith LLP, Haynes and Boone LLP, Greenberg Traurig LLP, Stikeman Elliott, FisherBroyles LLP, and Lowenstein Sandler LLP. Each of these firms handles relatively few filings, suggesting targeted use for specific technologies, jurisdictional coverage, or older matters connected to past projects.