Technical information

Frequently asked questions

1. Identification of the problem:
Contact us with a problem or an idea. We will accommodate suggestions for a process where we together identify the project needs and objectives.

2. Analysis and integration with existing IT systems:
Existing projects have already invested in infrastructure and systems. New project groups often have technology preferences. Together we analyze the systems and platforms that the Bionic modules will interact with.

3. Product configuration:
We typically deliver a combination of different modules configured to receive customer data and solve problems according to requirement specification. Together we develop reference libraries and a tailored API that makes the software very easy to use.
Bionic's approach to pattern recognition uses unique algorithms that calculate a mathematical distance between focus and reference. Then we use machine learning to define the threshold value to identify a match.

In most cases we only need a small set of references to provide satisfactory results, while in other configurations, larger amounts of training data may be required, for example, in recognizing unique objects in physical space.
Software modules and solutions are available for licensing. This includes a tailor-made API. Other configurations are delivered as development projects.


Bionic AI Suite ™ is designed for multi-platform, to easily integrate into existing projects.

The software modules can be installed run on everything from cloud services to small microcontrollers.


Bionic AI Core can process data from many different sources simultaneously.

Feedback loops

Bionic works with a project model that provides frequent launch rates and quick learning.

We participate from pre-production to production and ensure delivery to the customer's own resources.


Bionic was started as "IT Utvikling AS" in 1998, and has worked with consultancy in a wide range of innovation projects, focusing on embedded systems and sensor technology.

The Bionic pattern recognition algorithms have been in development for over 10 years, and in 2015 the products were expanded with machine learning and made ready for the market.