Tell us about your project!
Robotics
Achieving 99% data annotation accuracy in a short timeframe for autonomous navigation robots
Training data for AI and machine learning in robotics
About case:
About case:
Industry and use case
Robotics, autonomous navigation
DATA
10,000 street images
Project duration
1,5 months
Industry and use case
Robotics, autonomous navigation
DATA
10,000 street images
Project duration
1,5 months
Challenge:
Challenge:
The goal was to train the autonomous robot to navigate real-world environments. The client wanted to annotate 10 thousand images obtained from the robot's RBG camera to train the neural network. The quality of data labelling was paramount because the robot was going to navigate the streets of a metropolitan centre, and the cost of navigation error was high. It led to a large number of annotation classes.
Solution:
Solution:
The Training Data team consisted of 70 people and was split into 5 groups, including a separate quality assurance unit. Each group was responsible for specific annotation classes and verifying other teams' work. Teams organization in this structure allowed us to eliminate human annotation errors within our processes.
Outcomes:
Outcomes:
The customer trained a neural network and conducted field tests that demonstrated that the robot could select the perfect route to overcome obstacles.
99%
The customer saved a lot of time delegating quality assurance to our team
annotation quality
Feedback
Our goal was to pass the autonomous robot tests successfully. Thanks to the cohesive and independent work of the Training Data teams, we trained the robot on an excellent dataset and passed field trial.
Ivan N. Project Manager