Our Technology

Discover the technology at the heart of SLAMcore and how our expertise is helping to shape the future.

Spatial AI is the process of turning sensor information into actionable, spatial understanding.

Spatial AI is the process of turning sensor information into actionable, spatial understanding.  This concept is at the heart of what we do at SLAMcore.  The market opportunity for devices boasting reliable spatial understanding is vast. But high-performance solutions are out of reach for most OEMs.

Regular system failures in real world environments have become normalised. Solutions that work well inside the lab fail on the outside. SLAMcore is pioneering easy to use, high-performance solutions to bridge the chasm between demo videos and real-world performance, turning your product vision into a reality.

Robust, reliable performance with affordable hardware is within reach.

What is SLAM?

Among the many talented people at SLAMcore HQ, you’ll find some of the world’s most renowned experts in SLAM.​

To be truly useful, hardware like autonomous robots and high-end drones need to localise themselves, understand unfamiliar surroundings, and navigate with consistent reliability. This general problem of simultaneous localisation and mapping is best known as SLAM.

Among the many talented people at SLAMcore HQ, you’ll find some of the world’s most renowned experts in SLAM.​

We’re pioneers in:

  • Sparse feature-based SLAM for robots and drones
  • Dense SLAM systems using both monochrome and/or RGB-D cameras
  • High quality, tightly coupled visual-inertial fusion
  • Semantic mapping
  • Deep Learning for SLAM

Sensor Fusion

We are experts in the design and deployment of multi-sensor solutions that will provide you with the performance you require in the environments that matter.

Single sensor solutions will only ever get you so far.  Our algorithms are designed to seamlessly fuse information from a range of sensors to create a probabilistic model of the immediate world. That model provides your robot or drone with an answer to the SLAM problem that’s both accurate and extremely robust. We are experts in the selection, design and deployment of multi-sensor solutions that will provide you with the performance you require in the environments that matter.

Need proof? OKVIS, the open-source, state-of the art visual-inertial system available under BSD License was developed by SLAMcore co-founder Dr Stefan Leutenegger alongside experts at ETH Zurich. Just one example of our team’s experience developing widely adopted visual-inertial technology.

Deep Learning for SLAM

We are complementing the best from the old with the advances of the new.

Deep learning is revolutionising approaches in many technical disciplines and SLAM is no exception.  Although 100% “learned” SLAM solutions are still a long way from reality, here at SLAMcore we are complementing the best from the old with the advances of the new.   Our expertise includes:

  • 3D Object segmentation
  • Depth from video
  • SLAM in dynamic environments