Home  /  About & contact

Machine learning,
out of the lab.

AI Lab ApS is a specialist engineering and consultancy firm that brings Artificial Intelligence and Computer Vision to life in complex, real-world environments. Founded by five Engineers with a shared background in applied research, we specialize in moving machine learning beyond the laboratory and into rugged, “uncontrolled” applications.


The team

AI Lab is led by five specialists in computer vision, machine learning,
and sensor systems:

Anders Krogh Mortensen

Anders Krogh Mortensen

PhD · Co-founder

Takes vision systems from prototype to production: LiDAR and camera-based object detection, drone imagery pipelines and multispectral sensing. Tech lead on industrial vision-AI projects, with a background as Assistant Professor at Aarhus University.

Mads Dyrmann

Mads Dyrmann

PhD · Co-founder

Builds camera systems and machine-learning models that work outdoors in real-world conditions, from hardware prototype to deployed model. Advises industrial R&D teams on computer vision and AI. Former Associate Professor at Aarhus University.

Mikkel Fly Kragh

Mikkel Fly Kragh

PhD · Co-founder

Brings deep learning from research into production. His embryo-assessment models are used daily by hundreds of fertility clinics, and he works on driver-assistance vision and industrial quality inspection. Strong in MLOps, model validation and calibration.

Peter Hviid Christiansen

Peter Hviid Christiansen

PhD · Co-founder

Specialist in deploying deep learning on real hardware, from GPUs to microcontrollers. Experience includes active learning for self-driving cars, visual navigation for underwater vehicles, and end-to-end platforms for training and benchmarking vision models.

Søren Kelstrup Skovsen

Søren Kelstrup Skovsen

PhD · Co-founder

Develops complete vision systems, from hardware design and assembly to data analysis and model interpretation. Tech lead on vision-AI projects, with research experience in real-time computer vision for precision plant detection.

Our expertise

From sensor to trained model.

Good vision AI needs more than a good model. We cover the whole chain, from the sensor that captures the data to the network that learns from it.

01

Sensor systems

One of our core strength lies in developing perception systems that function reliably in “uncontrolled environments” characterized by changing light, unpredictable weather, and complex surroundings across industries ranging from manufacturing to infrastructure and nature.

02

Data Infrastructure

AI is only as good as its data. We build the pipelines necessary to capture, process, and store massive datasets from the field. From edge-to-cloud synchronization to structured data lakes, we ensure your data is accessible, high-quality, and ready for training.

03

Model Development & Training

We design and train custom Deep Learning models tailored to specific industrial needs. Whether it’s object detection, semantic segmentation, or anomaly detection, we optimize architectures for both accuracy and performance.

While our roots are in high-precision fields like agricultural engineering and biodiversity monitoring, our technologies and methodologies are industry-agnostic, serving clients across manufacturing, infrastructure, and automotive industry.

What we do

The full lifecycle
of AI-driven
vision systems.

We specialize in the full lifecycle of AI-driven vision systems:

01

Prototype Development

We take concepts from initial idea to field-ready hardware. We handle the entire stack: mechanical sensor integration, local processing software, and the back-end data architecture.

Cases →

02

High-End Consultancy

We provide specialist sparring for R&D teams facing complex challenges in deep learning and digital signal processing. We help navigate the path from initial feasibility studies to production-scale infrastructure.

Services →

03

Specialized Vision Systems

We offer mature, off-the-shelf products born from our internal research, such as the CamAlien (for invasive species and litter detection) and the CropCareCam (for automated crop and weed monitoring).

Products →

Research-active

Research-Driven Innovation

AI Lab remains a research-active organization. We continue to contribute to international journals and collaborate with leading academic institutions. By maintaining our own open-source datasets and publishing our findings, we ensure our clients always benefit from the latest breakthroughs in computer vision.

Recent publications →  ·  Open datasets →

Contact us

Let’s talk about
your system.

Feel free to get in touch if you have a question, want to build something together, or need expert help on your own project.

Address

AI Lab ApS
Rosbjergvej 22L
8220 Brabrand
Denmark

VAT: DK39858185