We are a robust AI-driven data management platform that solves complex data challenges for large institutions operating in dynamic and mission-critical environments with rapidly increasing volumes of data being ingested from a similarly rapidly growing number of data sources.
We solve these challenges by structuring “dark” or unstructured data at the edge, the location at which the data is generated and collected, and leveraging purpose-built AI models. Unstructured, or “dark” data, which is typically categorized as qualitative data, cannot be processed and analyzed via conventional data tools and methods. Conversely, structured data, typically categorized as quantitative data, is highly organized and easily decipherable by machine learning algorithms.
Structuring and then analyzing data using AI models at the edge, versus transmitting the data from the edge back to a central processing location for structuring and analysis, enables real-time decision making and data-driven operational efficiency.
We specialize in ingesting all available metadata from edge-based sensors used by government and law enforcement agencies around the world, including surveillance cameras (video), audio, telemetry, acoustic, seismic, and autonomous devices, along with large commercial corporations with fundamentally similar capabilities and requirements.
Data generated by these edge-based sensors, including video, can then be run through our trained AI models to detect objects present within the video frame. Once an object is detected, for example an automobile, additional identifying characteristics of the object can be extracted from the image including the license plate characters and the make, model, and color of the automobile. This process of analyzing, logging and categorizing ingested data is referred to as “structuring” the data.
Airship AI’s software allows customers to view structured data both in real-time as well as to conduct searches on the structured data at a later point in time. Real-time structured data use includes, for example, alarms on a specific license plate or a specific make, model or color of automobile. Non-real-time structured data use includes, for example, searching a database of video data that has been previously ingested and stored to find instances of a particular license plate being visible, along with other logged vehicle characteristics such as make, model and color of an automobile.
Additional edge deployed AI models enable similar object detection and recognition of common and custom trained objects, such as an aircraft, boat, person, animal, bag, or weapon. Airship AI’s models provide similar data points for these object types allowing analysts the ability to be notified in real-time of the detection of a specified object and similarly search for historically detected objects. Examples include detecting aircrafts and boats along with their respective tail numbers and hull registration numbers.
Our AI modelling process starts with pre-trained AI models from our technology ecosystem partners which we then customize using proprietary datasets tailored towards our customers unique workflow requirements. Where customers have pre-existing AI models or engines, we integrate those models or engines into our edge platform allowing customers to leverage proprietary models within the Airship AI software ecosystem.
Our primary offerings include Outpost AI, Acropolis, and Airship Command. Our offerings allow customers to manage their data across the full data lifecycle, when and where they need it, using a highly secure permissioned based architecture.
For more information, visit https://airship.ai.