Berlin-based AI company, Teraki, today announced it will enable efficient, mass volume collection of sensor data from cars on Microsoft’s cloud platform Azure, including the Connected Vehicle Platform.
Teraki (www.teraki.com) is the supplier of Intelligent Edge Processing for the automotive industry, building embedded software for highly efficient and accurate data pre-processing. The AI-based solution enables car makers to quickly process vast amounts of sensor data within cars in a highly efficient way. The efficiency gains go hand-in-hand with a strong compatibility with AI; the pre-processed data ensures more accurate predictions and detections of events, enabling customers’ algorithms to make better decisions. Intelligent Edge Processing is available for all telematics, video and 3D point cloud data.
Teraki Intelligent Edge Processing embeds on one side in the CPU & RAM-constrained chipsets of cars and integrates on the other side at a customer’s backend for the storage, labelling, training and analysis of data. Today, Teraki announces to build on Microsoft’s cloud platform Azure and the Microsoft Connected Vehicle Platform. The goal is to create leading end-to-end solutions for high volume data processing and analysis - targeting car makers.
High amounts of car data are required to train and develop new AI-based models. Therefore, it is important for OEM’s to train models in the cloud fast and efficiently. Teraki’s edge processing capabilities helps by enabling 10X more data per test-drive to enter the cloud. This provides faster insights about the relevance of specific data streams for the training and model update process. This helps in gathering more, high resolution data quicker, which speeds up training of new models such as e.g. predictive maintenance or autonomous driving functionality. Teraki’s technology and its support through the Teraki developer center reduces the number of miles required to train and retrain accurate machine learning models within the required accuracies for fail-safe operation, therefore speed-up the roadmap towards safe cars.
Not only for training models in the cloud but also for managing data streams from high volumes of cars efficiently the Teraki solution is a key component. The company’s embedded software lowers the costs of transmission of data; lowers the long-term storage costs of large amounts of granular data; and guarantees no loss of information. The Intelligent Edge Processing provides higher accuracy rates of customers’ algorithms and neural networks. I.e. using Teraki’s pre-processed data produces less False Positives and more True Positives at 10x lower computing power at the edge than any standard technology. With the same computing power, the accuracies of resulting machine learning models are enhanced by 10-30% compared to standard pre-processing technology. As Teraki technology is deterministic it is certifiable and qualifies for safety related applications.
By collaborating with Microsoft, Teraki enables the ingestion of more information to the cloud while lowering the cost for its automotive customers. This overcomes the challenges of collecting, transmitting, labelling and storing the enormous amounts of data streams from the hundreds of sensors inside cars. Creating AI-based models is key to the automotive industry to better manage fleets, improve cars and introduce autonomous driving functionality opening up the edge-of-the-edge within cars and enable safer cars.
“We are proud to be able to show a fully working data chain stretching from the individual sensors in a car up to Azure used by many of our customers in the field,” says Daniel Richart, CEO of Teraki. “Bridging the gap between the very decentral, CPU-constrained edges and the cloud is a key requirement of the industry,” he continues. “Like Teraki is leading in processing of edge data, Microsoft is leading with its cloud and edge platform approach. Bringing these two together, is a strong and logical combination for our customers.”
Tara Prakriya, Partner Group Program Manager Mobility & Connected Vehicle Platform at Microsoft, said, “Data pre-processing in the car is one of the big challenges in the automotive industry. Teraki will build solutions for efficient data handling on top of our Microsoft Connected Vehicle Platform. This enables our customers to build automated driving functions and customer services more efficiently.”
Teraki provides breakthrough edge data processing software to meet the exploding data demands of the $395 billion automotive electronics industry. The company’s AI-based Intelligent Signal Processing software delivers a more than 10X increase in automotive chip, communications and learning performance. This makes highly accurate AI applications possible at scale in embedded environments.
These leaps in performance enable the automotive industry to develop new, innovative and cost-effective ways to use the vast amount of data generated by in-vehicle sensors and control units (ECUs, MCUs, TCUs) to improve vehicle safety and autonomy at lower operational costs. Teraki has completed several pre-production validations by premium automotive manufacturers, as well as successful integrations on a variety of microcontrollers.
Headquartered in Berlin, Teraki is privately held and funded.