RocketHuntSmart job search
Reach out directly about this role
By city
3-5 years
Experience
Full-time
Employment
Hybrid, Onsite
Work Format
Middle
Grade
Data Analytics
Specialization
Robotics
Industry
Corporation
Company Type
Analyst-Developer for the Perception Annotation Team
Key characteristics of an autonomous vehicle are safety, comfort, and efficiency. Deep and systematic analysis of technology performance quality is a necessary step toward their widespread adoption and acceptance.
Quality analysis specialists work closely with research and engineering teams at all stages of software development. Their task is to ensure that the autopilot complies with high safety standards, current regulations, and the real demands of traffic.
Perception is the sensing system in autonomous transport. It processes data from sensors (lidar, cameras, radars) and forms a holistic understanding of the environment. This is the foundation for planning the behavior of an autonomous vehicle: you cannot avoid an obstacle if the system hasn't recognized it. The safety and development of autonomous transport directly depend on the completeness, accuracy, and quality of the description of the surrounding world.
Working with autonomous transport perception system data You will need to understand how autonomous systems 'see' the world: what lidar is, how it works, what data it provides, and what limitations it has. A significant part of the work will also involve analyzing data from cameras.
Building the scene: 3D representation of surrounding objects Data from all sensors is combined into a unified spatial representation of the environment. At this stage, computer vision methods are used to fuse data from different sensors into a single scene representation.
Participation in solving practical tasks of autonomous transport Based on the collected data, you will participate in solving applied tasks that directly affect the safety and correct behavior of the vehicle, including: * detection of road defects (e.g., potholes) for timely reaction; * recognition of various traffic light types and determination of signal changes; * detection of special vehicles (e.g., ambulances) and correct interaction with them.
These are not abstract tasks: we see how the results of our work change the behavior of autonomous transport in real road conditions.
More about analytics at Yandex — in the channel Yandex for Analytics