AI cameras, or computational photography, is a staple of the modern smartphone market. Nearly every smartphone uses some form of AI to help the pictures look better. The future of photography will be more computational than optical of nature.
About Us.
What. How. Why.What we do.
Teraki’s software is a key driver for scaling of Insurance, Predictive Maintenance and Autonomous Driving Applications by enabling access to more qualitative data.
Teraki begins doing so, by identifying the most relevant data already at the embedded side and transmits this data in the proven most compact form possible. This is done while maintaining the customers key KPIs based on the latest state of the art AI data processing schemes. As such, Teraki’s software enables to fulfil low-latency SLAs required by the end-user and ultimately to adapt key KPIs such as cost-per-mile at the highest precision and lowest latency.
Therefore, Teraki enables to bridge the gap towards true vehicle autonomy, by overcoming the time-gap of future 5G networks and by optimising automotive hardware cost efficiency.
How it works?
The data reduction is achieved by pure software. In the car (sensor, microcontroller, gateway, etc.), a client software with lowest requirements is deployed and identifies the data streams required by the AI based processing scheme. Only a fraction of the data is effectively acquired and encoded.
The data is recomposed and fed into the customers AI processing schemes, relying either on the fully reconstructed or reduced data. No relevant information is lost, while maintaining between x4 to x15 more data rates than accessible with compression products.
With a proper analysis of the data in a telematics device, higher reduction rates are achieved. Teraki’s products are compatible with existing compression and AI chip based processing solutions.
Key features.
Compared with alternatives:
Your CPU processes 5x more data.
Your application runs with 15x less end-to-end latency.
Your mobile data consumption is up to 50x lower.
Key insights are gained already at the device level while maintaining lower hardware requirements than NN based solutions.
100% software (client - server architecture). No additional hardware required.
It fits in constrained environments, the client component is just a few KB small.
Works with any private or public cloud infrastructure.
Suited for real-time applications.
Compatible with all protocols, networks, encryption and hardware technologies.
Compatibility with latest hardware platforms including AI based processing chipsets.
Use Cases.
Examples of mobility applications that Teraki makes happen.
Automotive Insurance
Use Case: Crash detection
Crashes form the largest category of costs to insurers.
Better and more complete crash detection significantly improves the main KPIs for insurance companies as with more available data algorithms can achieve higher accuracies and draw more relevant conclusions.
Within existing architectures and hardware Teraki technology delivers 30% higher accuracy rates on crash detection algorithms. In addition, Teraki gets you this data 10 times quicker, thereby improving of the key KPIs for crash detection: accuracy and latency
Find out morePredictive Maintenance
More data leads to less down time & costs and to higher customer satisfaction.Getting more data out of a car with existing hardware (e.g. microcontroller, gateway, telematics unit) enables to do good analysis to determine when maintenance is due.
Find out more
Autonomous Driving
Autonomous Driving is highly dependent from new sensors such as lidar, radar, video, etc. These sensors have in common to exponentially produce more data than the existing telematic sensors. This increases the data processing even more.
Find out moreBenefits.
Core benefits our solution provides.Jobs.
Working at TERAKIView our job page to see what position we have currently open.
We are a diverse team of various backgrounds and are always looking for exceptional people. See open positionsPartner & Awards.
We received awards and are supported by:
Blog.
Latest Blog PostsThe sensors and sensor processors are by far the highest contributors of power consumption of Autonomous Driving. Going autonomous demands more power that will be taken from battery. Three cars running autonomously for a year in can produce from 1.3 to 0.2 Tons CO2 per year depending on the spec.
Autonomous driving algorithms detect and label objects in streams of complementary types of data. In general it is easier to detect objects in point clouds, than it is to identify correctly their type. An image can identify easily what object is in it, but it is more complicated to build a bounding box around the object