Space mission’s scenario is rapidly evolving. However, the transition from a traditional space model into a commercial one is already showing some bottlenecks and barriers in specific applications preventing new market opportunities to flourish and limiting the effectiveness of the services delivered to ground.
In several verticals where EO data is today used (e.g. agriculture), users/customers are being focused on their core business process, and less interested in data itself and more in applications to make their business more efficient and effective.
In some context (e.g. fire detection) the right information shall be provided to end users/customers at the right time and in the right place. And the place can be in some cases the space segment, where the availability of actionable information can be a game changer. In this approach part of the EO value chain is moved from ground to space to promptly transform sensed data into “wisdom”, to exploit it directly or to enable an optimized exploitation of limited resources on-board (in fire detection this means downlinking just the warning with its geographical location, few tens of bytes instead of raw GBs). We defined this approach as “SpaceStream”.
Although some of the barriers are currently being addressed (e.g. launches availability and satellite deployment flexibility, versatile ground infrastructures and software, suitable regulations, etc.), existing satellite operators, incoming new satellite operators and end users of space data are still experiencing severe inefficiencies affecting traditional mission operational models:
• delays in decision-making due to ground operator intervention,
• missed observation opportunities due to limited on-board autonomy (data analysis is currently largely executed on ground),
• missed information due to limited downloadable data from satellites (small satellites usually offer limited power, bandwidth and controllability),
• poor quality or irrelevant downlinked data,
• late systems failures detection.
Autonomous operations and Artificial Intelligence on-board are key enabling technologies able to impact over these limitations. Main expected impacts will affect reactivity, responsiveness and latency.
The advanced functionalities based on detection, extraction and exploitation of data information content will be the core value points in this paradigm shift, moving focus beyond “raw” data. Most of these new functionalities will have to manage huge amounts of raw data, implement structured and automated information refinement processes, able to react in real or near real time to single pieces of information, to touch off new tasking for improved EO acquisitions specifically targeting the detected needs.

In order to provide such solutions to the highlighted problems, AI-express (AIx) aims at building a new AI enabled processing framework, embarking it on satellites and then bringing into the market a new concept of satellite as-a-service. AIx will make available in-orbit resources on-demand: EO data and above all, actionable information and actions at the right time.
AIx is a set of services à la carte provided by a flying satellite and based on a public catalogue with an “app store” approach. Services will include EO data acquisition, processing (actionable info extraction), downlink and distribution. Ready-made applications (e.g. a fire detection and warning service) will be made available on the app-store and services can be also combined together to build custom applications.
Useful information can be then transferred back to Earth as notifications and alerts, or directly exploited on-board in autonomous decisions workflows. Useless data and information can be completely deleted without loading any precious resource like memory and bandwidth.
At regime AIx services will be deployed on a fleet of AI enabled satellites (D-Orbit’s ION enhanced platforms) and will be available to any user customer interested in testing applications based on AIx EO payloads’ data on-board (and also to third-party payloads embarked on IONs). Commercial services will be based on the pay-per-use usage of the on-board assets that will be negotiated directly among on-board sub-systems and accounted thanks to blockchain type mechanisms.

What distinguishes AI-express from the rest of the market competition is the capability to provide a set of deeply configurable services, scaling from pay-per-use to full missions as-a-service. Services allow for the evaluation of new approaches to space missions and for the validation of novel concepts in the real environment. Flexibility and configurability allow for iterating in-orbit tests and fine tuning of applications.