AI for Earth and everything on it Easily deploy planetary scale AI on physical world data The only integrated service for building intelligence with vector, raster, and tabular data together, at any scale. TRY WHEROBOTS Integrated AI for Earth Models Built-in models for your use case, or bring your own. Start in seconds Evaluate AI for Earth models instantly with a few lines of code. Cost-effective at any scale Analyze a single region or the entire planet, only pay for what your use. Unlock Geospatial Intelligence with AI See how state of the art AI for Earth models can tackle your real-world challenges. AI for remote sensing imagery Map matching Raster and vector analysis Automate and scale Apply AI models to remote sensing imagery Use state of the art AI models on remote sensing imagery for object detection, segmentation and classification. Or bring your own model. Learn More Improve mobility data quality with map matching Map matching snaps noisy GPS data to road segments with confidence, simplicity, at high performance and scale. Learn More Combine raster and vector analysis Seamlessly store model inference results in Wherobots for further analysis and enrichment with vector or tabular data. All within a single platform and single development experience through Python or Spatial SQL. Learn More Automate and scale Scale your analysis from small-scale prototyping to planetary-scale production workflows, only pay for what you use. Learn More Benefits of AI for Earth Built-in models or bring your own Pre-integrated TorchGEO models for segmentation, classification and object detection, or bring your own PyTorch model. Learn more Expertise included Accelerate your time to insight. Pre-configured data sets, transforms and models enable anyone to start using GeoAI models today. Learn More Scalable model inference Scale your analysis from a single region to the entire planet, seamlessly and cost effectively. Learn More Integrated with WherobotsDB Post-process, analyze or join the model outputs with business context, using Wherobots’ scalable spatial data processing engine. Learn More Cost-effective The most cost-effective option for small and planetary-scale analysis. Only pay for what you use. Learn More Instant results Enable any developer to easily evaluate AI for Earth models with a few lines of code. Learn More Put your spatial data to work with AI Wherobots AI for Earth eliminates time intensive, manual processing of imagery, so you can spend more time using insights and data vs manually scanning and classifying context in your datasets. LEARN MORE Model Output Examples Sample insights from models supported by Wherobots Airplanes detected with SAM2/OWLv2 Find airplanes from open satellite imagery using Wherobots and the SAM2 and OWLv2 models from a single text prompt. Solar farms detected with raster inference Detect solar farms in open satellite imagery using Wherobots and the Allen Institute Satlas models. Loading satellite imagery with STAC reader Efficiently work with satellite imagery datasets by directly loading STAC items and collections into Sedona DataFrames with the built in STAC reader. STAC reader COGs MLM Learn More df = (sedona.read .format("stac") .load("https://earth-search.com/sentinel-2-c1-l2a")) Inference with SQL and Python Run model inference with a few lines of SQL or Python code and then post-process or join the results with WherobotsDB. SQL AI Remote Sensing Python Segmentation Learn More -- Find bounding boxes for 'solar panels' CREATE OR REPLACE TEMP VIEW detected_bboxes AS SELECT raster_column, RS_TEXT_TO_BBOXES('SAM 2 ', raster_column, 'solar panels') AS detection_result FROM my_imagery_table; Bring your own model Register your own PyTorch model with Wherobots and instantly generate insights at scale. MLM Custom ML Models Learn More user_mlm_uri = [PATH-TO-YOUR-MLM-JSON] predictions_df = sedona.sql(f""" SELECT rast, segment_result.* FROM ( SELECT rast, RS_SEGMENT('{user_mlm_uri}', rast) AS segment_result FROM df_raster_input ) AS segment_fields """) FAQ What types of computer vision tasks can I perform with Wherobots? Wherobots supports the primary computer vision tasks required for geospatial analysis. This includes Image Classification (identifying a single class for an entire scene), Object Detection (locating and labeling specific objects), and Semantic Segmentation (classifying every pixel in the image). How do I use my own model in Wherobots? To use your own model with Wherobots, you first convert your custom PyTorch model into an inference-ready format, such as Torchscript or a PT2 Archive, and upload it to your Wherobots Cloud-accessible storage. Next, create and upload a descriptive MLM JSON file that uses the Machine Learning Model Extension Specification to define your model’s properties. You can then execute the inference on raster data by referencing the uploaded model’s MLM URI within a GPU-Optimized Wherobots Notebook or Job Run. For more information, see the documentation. What are some common real-world use cases for Raster Inference? Raster Inference can be used across many industries to extract meaningful insights from large-scale remote sensing imagery. Typical use cases include: Agriculture (monitoring land use and predicting crop yields), Environment and Conservation (tracking biodiversity changes and detecting oil spills/deforestation), Energy (assessing land suitability for renewable projects), and Map Creation (extracting features like buildings and road networks). Get started Modernize your spatial data in the lakehouse today. request demo try wherobots now
