NorCom selects MapR to accelerate deep learning in autonomous driving applications

MapR helps organisations scale machine learning applications using distributed computing

MapR Technologies has announced that NorCom, a full-chain supplier for big data solutions, has selected the MapR Converged Data Platform to serve as a foundation for its autonomous driving applications that leverage deep learning technologies. The partnership enables joint customers to deploy deep learning frameworks that can provide fast and reliable analysis in critical compute environments.

"The automotive industry is undergoing a massive disruption at the moment, fueled by advanced technologies such as deep learning," said Dr. Tobias Abthoff, member of the executive board, NorCom. "We help German automotive giants Audi and Daimler harness the power of deep learning in order to achieve their goal of building autonomous vehicles. In MapR, we found the perfect platform to scale our deep learning efforts using distributed storage and compute."

NorCom leverages a purpose-built deep learning framework for the automotive industry. To fully take advantage of it, they required a way to efficiently manage the massive data sets generated by sensors and cameras in self-driving cars. Running containerised deep learning applications on the MapR Platform provided the required speed, scale and reliability to successfully analyse continuous data in an autonomous driving environment and achieve the benefits of deep learning.

"We are proud to collaborate with industry leaders like NorCom who are transforming the way data is used," said Anoop Dawar, vice president, product marketing, MapR Technologies. "Autonomous driving applications are a perfect example of how a scalable and reliable data platform can successfully enable deep learning in a real-time production environment. Volumes of data can quickly outgrow traditional data management approaches. The MapR Platform provides a modern and flexible foundation for deploying applications that help organizations innovate, compete and thrive."

NorCom also selected MapR as a platform because it helps support machine learning applications in Docker containers, which helps to scale the applications with agility. The MapR Converged Data Platform works with multiple deep learning frameworks, including TensorFlow and Caffe, providing organisations the flexibility to choose the best suited framework for their use case. As a single platform across all data centres, cloud deployments, and edge clusters, the MapR Platform makes it easier to roll out machine learning models so that they can be applied to newly created data anywhere across the data fabric. Moreover, the newly created data can be immediately added to the training data set because both reside on the same platform. This approach is designed to simplify the goal of continuously improving the accuracy of machine learning models.

NorCom also supports real-time processing of the MapR platform to support real-time processing for a broad variety of data types, including files, documents, images, database tables and data streams. The MapR platform is engineered to meet the most stringent speed, scale and reliability requirements within and across multiple edge, on-premises and cloud environments. The platform is designed to serve as a system of record and to support transactional workloads, with comprehensive high-availability, data protection and disaster recovery capabilities.