Cyber Cognitive Intelligence (CCI)

The Department Cyber Cognitive Intelligence (CCI) assists companies, especially SMEs, in introducing and using artificial intelligence (AI) and machine learning (ML) methods. To this end, it offers the complete range of implementation services, from identifying and roadmapping use cases for AI in companies (AI Explorer), to rapid feasibility analyses (Quick Check) and operationalization (AI Services), as well as auditing AI solutions (AI Audit).

On the research side, the department focuses on dependable, explainable, robust, and data-efficient learning methods, the development of intelligent planning and optimization algorithms, and on coupling quantum computing with machine learning. The research results are used in many branches of industry, such as the machinery and equipment sector, automotive, financial services, process industry, and life sciences.

The department's successful application solutions include intelligent algorithms for production processes and job scheduling, audited algorithms for fraud detection, methods for explaining time series data and image data, and self-learning AI control algorithms.

Our business areas and services

 

Dependable AI

Artificial intelligence delivers high performance in many application areas. Nevertheless, it is rarely used in critical applications (e.g., in the medical industry) due to its limited interpretability. To change this, one of the groups in the department focuses on methods for explaining and verifying AI systems, as well as for quantifying uncertainty.

 

Autonomous Planning and Decision-making

Artificial intelligence methods are ideally suited for modeling situations in planning and decision-making processes that would be too complex to model using traditional techniques. The focus here is on incorporating uncertainty, domain knowledge, and foresight in order to generate robust autonomous planning and decision-making systems.

 

Data-efficient AI

Vast amounts of data are generally required to utilize machine learning. However, large data sets are not always available, and data collection and annotation are invariably expensive and time-consuming. Research on data-efficient AI therefore looks at how this problem can be overcome and how effective machine learning models can be implemented even in cases where real data is sparse.

 

Quantum Computing

Quantum computing holds the potential to efficiently deal with problems in the future that are unsolvable today. With the two focal points quantum machine learning and quantum optimization, the department studies industry-relevant applications and methods in order to support the transfer of quantum computing on a broad industrial scale at an early stage.

 

Cooperation formats

The Department Cyber Cognitive Intelligence (CCI) supports manufacturing companies and their suppliers across a wide range of innovation topics from the initial idea to the final implementation. The range of services includes feasibility studies, quick checks and workshops, as well as the development of complex technical manufacturing modules utilizing machine learning. Project formats range from short-term, compact developments to specific assignments and long-term strategic cooperation.

 

AI Innovation Center

The AI Innovation Center “Learning Systems and Cognitive Robotics”, which forms part of the Cyber Valley AI research consortium, helps companies take advantage of the economic opportunities offered by artificial intelligence and especially by machine learning.

 

TRUMPF Lab

Improving data quality is the most important step towards a maximally automated production. With TRUMPF, the department is pursuing AI-based order tracking in order to identify potential bottlenecks in production and aid skilled workers.

 

ELISE

ELISE is the “European Network of AI Excellence Centers” that cooperates closely with ELLIS (“European Laboratory for Learning and Intelligent Systems”). The goal is to spread knowledge and methods among science, industry, and society.

 

AI-based production planning and control

Together with Porsche AG, Fraunhofer IPA is developing an AI-based production planning and control system. The aim is to anticipate customer wishes at an early stage and incorporate them in the order planning process. This will make planning more robust, especially in times where supply chains are volatile.

 

Transaction Miner audit with Experian

As an independent partner, Fraunhofer IPA reviewed the AI component of a product on behalf of the company Experian.

 

ML4SAFETY

In the ML4SAFETY project, an integrated framework for verifying “safeML” is being developed, which will allow innovative machine learning methods to be used in applications where safety plays a major role. This will help manufacturers and suppliers of autonomous, safety-critical systems as well as testing and approval institutions to bring demonstrably safe ML-based systems onto the market.

 

AutoQML

Automated machine learning (AutoML) enables low-threshold access to AI solutions. The AutoQML project extends this approach by quantum computing-based methods to speed up the transfer of this technology to industry.

Press release March 2021

New AI applications for SMEs in Baden-Württemberg

AI expert Prof. Dr. Marco Huber in discussion

Science meets Fiction – AI in the SWR-Thriller 'Exit'

Press release November 2020

Fraunhofer IPA and HLRS start cooperation

Press release Ocotber 2020

AI makes it possible: Tomorrow's customer wishes, already planned today

Press release September 2020

Thanks to AI, robots are learning how to perform assembly tasks

Video

Keynote by Marco Huber at Nexcon

Source: Staufen.Digital Neonex/NEXCON

Podcast

“AI in Industry”

In the “Factory of Tomorrow” series, our AI experts present scenarios for using artificial intelligence in automation as well as current projects worked on by Fraunhofer IPA.

Annual report 2019, P. 21

AI is expected to induce a growth spurt four times higher than that of the steam engine

Book reviews on Artificial Intelligence and Machine Learning

Daten