Avelabs

Yonohub: A System for Developing Autonomous Vehicles and AI in the Cloud


Autonomous Vehicles, especially in the era of AI, include thousands of algorithms developed using hundreds of technologies and trained on Big Data. This poses plenty of ever-growing challenges to develop, benchmark, reuse, and integrate these algorithms. Yet, researchers and engineers are still tackling these challenges using inadequate development tools and workflows from the past. Yonohub is the first system to tackle these challenges by moving this entire workflow from PCs to the cloud. This new approach opens new horizons for modularity, reproducibility, collaboration, and integration. With a block-based design inspired by AUTOSAR and ROS, Yonohub encapsulates anything in a block: an algorithm, a dataset, a simulator, a sensor, or anything else. Yonohub integrates blocks developed in Python, C++, MATLAB, ROS, and adaptive AUTOSAR applications. Once a block is developed, it is shareable and reusable by others with just drag-and-drop. A block can be deployed in a car on the road while the rest of the pipeline is running in the cloud. This is useful for data acquisition, MIL, HIL, quick hardware deployments, etc. To bridge the gap between researchers and companies, Yonohub includes a marketplace in which everyone can publicly share and monetize their blocks, environments, and datasets, while still protecting everyone’s intellectual property. There are also several other apps for building custom runtime environments, scheduling long-running jobs, storing huge amounts of data, source code management, and a web IDE.

Yonohub is available Online under (www.yonohub.com) and On-Premises.