A simulation platform to train AI to understand the world.

Our Demos

Our synthetic data and vision APIs can significantly improve machine learning algorithms and enhance AI applications in industries by providing the scale and diversity.



Manage Construction Sites


  • Improve safety and productivity of site by the understanding the activity of workers and machines on location.

What We Did

  • Environment: rural terrain based on OSM, DEM height maps, satellite data and GPS coordinates or real world location

  • Objects: machinery, equipment, materials, people

  • Annotations: 2D bounding boxes and semantic segmentation





Ensure Public Safety


  • Improve response time of first responders by detecting unusual activity in urban settings

What We Did

  • Environment: Urban setting based GPS coordinates of a real world location

  • Objects: Buildings, vehicles, people, street objects

  • Annotations: 2D bounding boxes & semantic segmentation





Maintain Machinery


  • Reduce hazardous tasks and labor costs associated with infrastructure inspection of machinery

What We Did

  • Environment: Rural based on GPS coordinates of real location

  • Objects: Machinery, buildings, equipment

  • Annotations: 2D bounding boxes  




Detect Weeds


  • Reduce the use of pesticides by identifying noxious weeds

What We Did

  • Environment: Rural based on GPS coordinates of real location

  • Objects: crops, weeds

  • Annotations: 2D bounding boxes, 3D bounding boxes, semantic segmentation



Case Study

Preserve Wildlife

“The final results from Great Elephant Census show 352,271 African savanna elephants in 18 countries, down 30% in seven years.” 

- The Great Elephant Census



The African elephant, the largest land mammal in the world, is a keystone species that plays an outsized role in maintaining the fragile ecosystem in which they live. Due to a booming ivory trade, African elephants in 18 countries have found their numbers decline from an estimated 3-5 million in the 1930s, to roughly 350K remaining today.  



One of the key challenges in this conservation is simply counting them and be able to track their numbers effectively.


What we did

  • Environment: Rural environment based on GPS coordinates, OSM, DEM, satellite data for Botswana, Africa

  • Objects: elephants

  • Annotations: 2D bounding boxes

  • Algorithm Used: YOLO-v2 (You Only Look Once)


VIDEO DEMO: Comparison of performance of an object detection algorithm trained on real data versus synthetic data


What We Learned


Test 1

Microsoft COCO Real Data


We took a set of pre-trained weights for this algorithm, (already trained on a large dataset called MS (Microsoft) COCO which contains data across 80 object classes including elephants), and tested it on this video to obtain results.


The algorithm constantly misidentifies an object as a bird, sheep or cow, but rarely an elephant.  It turns out that the MSCOCO dataset does not contain many images of elephants from an aerial perspective.


Test 2

AI.Reverie Synthetic Data


We further reviewed the data to understand why this might be the case. We then created another scenario where we train YOLO from scratch, but included our AI.Reverie synthetic data which contains aerial perspectives of elephants.


Our synthetic data of elephant images at different perspectives augmented the training of the algorithm and greatly enhanced its ability to detect elephants.


Tell us about your data needs.