Developer for Robot Delivery Motion Planning Group

At Yandex, we are taking a fresh approach to deliveries from stores and restaurants: we aim for orders to be delivered not only by humans but also by delivery robots. We want to make the delivery robot faster and more reliable. To bring this to life, we are building a team of active and passionate professionals.

What tasks await you

Safety Framework Development Robot motion planning can be solved algorithmically, covering all possible real-life situations with heuristics, or by leveraging the power of ML. Our colleagues are already working on training an ML planner, but we cannot fully trust it yet: it provides no guarantees, and its output needs validation. You will be responsible for developing, implementing, and testing the architecture and its components that will allow using the ML model's output with safety guarantees and the ability to switch to an algorithmic planner in exceptional situations.

Robot Speed Optimization To increase the happiness of delivery users, we need to make our robots move faster. This task includes both finding and eliminating problematic areas in the robot's behavior and accelerating existing algorithms within the constraints of its hardware.

Reducing Dependency on Detailed Maps We strive to reduce the amount of preparatory work when launching delivery robots in new locations. Therefore, for building routes and determining safe movement paths, we plan to use publicly available maps combined with the robot's own perception capabilities.

We expect that you

• Have built high-load systems with strict time and quality constraints
• Have an excellent understanding of classic data structures and algorithms
• Write code and are proficient in C++
• Are able to explain your decisions and defend them against opponents
• Take criticism calmly, know how to work in a team
• Are responsible and organized

Will be a plus

• Have solved motion planning tasks
• Are familiar with Python/Go