Digital Process Development in Mobility and Infrastructure

AI-powered solutions for data-driven value creation

Small and medium-sized enterprises are under pressure: a shortage of skilled workers, complex supply chains, increasing sustainability requirements – digital solutions are more in demand than ever. However, the introduction of new technologies such as artificial intelligence often fails due to a lack of resources, guidance, or trust.

We develop data-driven, AI-supported solutions that enable companies and public authorities to make processes more efficient, ensure quality, and manage resources in a targeted manner. We do not just rely on classic data-driven methods, but we use generative AI to derive knowledge from data and have created a new generator category in addition to publicly available text and image generators: the numerical transformer, which can predict precise probabilities.

We also go one step further: Agentic AI creates systems that are no longer mere tools, but can analyze, decide, and act independently. For the first time, a tool becomes an active player. This new class of intelligent agents takes automation to a whole new level – and opens up enormous economic potential. From simulation and predictive maintenance to real-time decision preparation, our solutions integrate seamlessly into complex applications and deliver added operational value.

With domain-specific models, powerful infrastructure, and an interdisciplinary team, we accompany you from initial idea to implementation – in a practical, forward-looking manner and with a view to sustainable value creation.

Discover the next stage in the evolution of artificial intelligence with us – and turn your data into an active value creation partner.

Our Expertise

© Oxy Maru | Adobe Stock
  • Optimized selction and configuration of high-performance GPUs for AI and data applications
  • Experience in NVIDIA CUDA and H100 architectures
  • Deelopment of GPU cluster management methods for the training and inference of  medium-sized and large models

© Gorodenkoff | Adobe Stock
  • Building of scalable data infrastructures (e. g., data lakes, warehouses, GraphDB hybrid architectures)
  • Real-time processing of large data volumes
  • Data quality, safety and governance (e. g., compliance with German data protection regulations DSGVO)
  • Integration of heterogeneous data sources and APIs

© The Little Hut | Adobe Stock
  • Development and fine-tuning of generative models (LLMs, diffusion models, etc.)
  • Ultra-pure transformer for numerical prediction
  • Applications in text and code generation
  • Prompt engineering and evaluation of generative output
  • Establishment of secure and robust AI pipelines for the manufacturing and service sectors

© Jay Koppelman | Adobe Stock
  • Design and implementation of autonomous AI agents
  • Multi-agent systems with planning, decision-making and interaction capabilities
  • Integration of tool use, API access and memory structures
  • Applications: Assistance systems, process automation, simulation, e. g.

Equipment and laboratories

GPU cluster for large and medium foundation models and transformers

  • Computing node​s featuring 8x NVIDIA H100 NVL
  • Storage node​s on 8x15TB NVMe​

Small GPU cluster for small and medium foundation models

  • Computing nodes featuring 1584 CPU cores, 5 TiB RAM and 8 Nvidia Cuda GPUs with 46GB RAM​ each
  • Storage nodes on 449 TByte HDD​

End-to-end solutions for industry

Generative planning of new and existing utilities supply in semiconductor factories

FastFlow2D

  • Currently under development
  • Digital support and partial automation of planning processes by applying algorithms and artificial intelligence (AI)
  • Faster and more precise planning decisions with the help of automated algorithms
  • Significant shortening of the process from six months to two weeks

Efficient search for new materials combinations based on generative materials engineering AI model

MATgix

  • AI-aided material development
  • End-to-end solution using high-performance infrastructure
  • Strategic advantage in case of market uncertainty

Smart Infrastructure Maintenance Analysis & Prediction

SIMAP

  • Predictive maintenance planning for infrastructure systems
  • Transferable model based on real city data
  • Practical AI wizard for administrative offices and infrastructure operators

Optimization of additive manufacturing processes through data-driven predictions and machine learning

CertiFlightAI

  • Automated defect detection in additive manufacturing
  • Real-time prediction of mechanical properties
  • Integrated data base for continuous quality assurance
 

TASC

 

  • Standardized data platform for simulation-ready traffic and accident data
  • Virtual testing scenarios from real traffic scenarios
  • Data-driven security of ADAS in early development stages

Platform for global ESG risk analytics

Terrisq

 

  • Location-specific ESG risk assessment on the basis of global data
  • Combination of semantic analysis with stakeholder and network data
  • Interactive GIS visualization for risk assessment purposes

Training and transfer

  • Industry-specific services, e.g.
    • Potentials of generative AI in economy and administration
    • High-performance AI in action
    • Strategic data management 
    • Agentic AI systems

Projects

 

Shared data space for the construction sector on the basis of Gaia-X

iECO

  • Digital twin for the entire building life cycle
  • Efficiency due to shared data infrastructure
  • Basis for digital testing and approval methods

Continuous digital quality assurance chains for innovative approval processes at the example of additive manufacturing

CertiFlight

  • AI-supported big data platform for the analysis and certification of additively manufactured components in aviation
  • Identification of parameters relevant to quality with the help of self-learning algorithms
  • Development of resource-friendly digital testing methods
  • Interdisciplinary consortium covering the entire value chain from  machine manufacturers to approval authorities to users

Automation of planning for factory construction and conversions

FastFab Design

  • Testing of AI-based methods that digitally support planning processes for the conversion of semiconductor factories
  • Development of data-driven decision-making models that improve planning processes through precise predictions and structured alternatives 
  • Shortening of typical planning periods through algorithmic procedures
 

Data-based mobility and energy services for the expansion of public charging infrastructure

DatenTanken

  • Digital business models for the operation of public charging infrastructure
  • Combining charging infrastructure with vehicle data for use in digital services or data monetarization 
  • Data-driven reduction of emissions and operationg costs

AI-powered platform for digital testing and approval procedures in additive manufacturing

AMCOCS

  • Data platform for the digital certification of additively manufactured components
  • AI-supported prognosis of component strength prior to the printing process
  • Digital business models for industrial application

Automated invoice posting detection

AuReBu

  • Self-learning AI platform for accounting processes in the tourism sector
  • Integration of machine learning in sector-specific workflows
  • High automation rates through continuous learning

Business models for autonomous driving

Cartox²

  • Development of digital business processes and services for automated driving on the basis of hybrid real-time data
  • Integration and analysis of car-2-car, network coverage, and routing data on a platform for the provision of low-latency mobility services

Intelligent Traveller Early Situation Awareness

iTESA

  • AI-based alarm system for detecting travel risks in real-time on the basis of social media and public data sources
  • Analysis of critical events with the help of semantic methods and historical travel data – potential for transfer into further sectors such as logistics

Internet Social Analyser in Natural Disasters

iSand

  • Automated real-time risk information system for disaster situations
  • Processing of information shared in social netzworks with methods of machine learning for the purpose of directing crowds to perform and receive organized, suitable support measures

Risik assessment for multi-modal transport chains

Synchro-Net

  • Self-learning assessment system for risks in multi-modal transport chains
  • Provision of risk indicators relevant for decision-making with the aim of optimizing logistics planning

Intelligent IT-supported platform for electric, sustainable and efficient infrastructure and fleet management in logistics hubs

iHUB

  • Predictive battery deployment management and dynamic route planning