Teraki is the leading supplier of edge processing software for the Automotive and IoT industries. The company’s technology makes applications and algorithms run highly efficiently, faster and at the highest accuracy.

• Connected Cars: New use cases made possible by retrieving data from cars and better understanding & managing individual cars and fleets of cars.
• Electrification: % of electric driven cars is increasing. This drives different competition and the need for efficient monitoring and matching of battery charging needs.
• New Mobility models: New models such as ride-sharing, ride-hailing, short-term rental, ride-bundling, etc. are changing the classical “ownership-model”.
• Self driving cars: Autonomous Driving (AD) will take some years, but will fundamentally change the industry and lower the cost per mile by 72%. ADAS already in production.

Many of the above opportunities are based on accurate AI-models. Accurate AI-models run on high volumes of data from car’s sensors.  The demand for high quality sensor data causes the below four challenges that Teraki helps to overcome:

The continuous data loop to train, build and update AI-models requires an ongoing ingestion of sensor data and selecting relevant ‘events’ for further training. Teraki has fully automated this edge processing part of this process via the DevCenter platform. Customers can – in an series production scale - easily trigger API-calls for the requested signals or events to further increase the accuracy of their AI-models.

• More accurate machine learning: Customers can continuously train and update pre-processing and their models in the car. Algorithm’ accuracies are increased with 10-30%.

• Lower costs: Customers lower their mobile network and storage costs for data that they need to send and/or store.

• 10x more efficient CPU utilization: Up to 10X more efficient utilization of available computing power resources additional to AI Chip optimization. Up to 6X more efficient RAM utilisation on top of AI Chip optimization.

• Quicker applications: Customers achieve lower latencies. 60% of application latency is data pre-processing and 40% is on ML. Teraki reduces this 60% with factor 10X.

AutoSAR: “The services of the Teraki library are always executed in the context of the calling function. There are no additional dependencies required, meaning that all AutoSAR versions are supported.”

OS: “The Teraki solution is packaged as a stateless library, meaning there are no special OS requirements (it can be integrated with any OS or simply within a bare metal environment).”

Security: “Device connectivity is not handled by the library itself, therefore the security requirements regarding the communication are handled by the user application. Teraki is agnostic to and compatible with other security encryptions.”

Teraki has competences of intelligent data pre-processing in highly constrained environments and do this in a deterministic way. AURIX™ is leader for automotive-grade chipsets for security-related applications and functions. Both form an unique combination of leading specialist in their respective fields and provide customer with a best-of-two-worlds solution.

Target applications:

• Data collection campaigns
• Predictive Maintenance
• Driving behavior
• Driver attention
• Crash detection
• Radar accuracy
• Electric charging
• Battery management
• PHEV efficiency
• ACC improvements
• Sensor Fusion
• Object detection
• Autonomous driving