Space AI application accessible with IBM by Ubotica

Space AI Ubotica Technologies

Leader in space AI Ubotica Technologies has partnered with IBM to use watsonx.ai components and IBM cloud infrastructure, with the goal of making it easier for developers to get their applications operating on satellites. Mutual customers will be able to safely deploy their AI models on satellites that make use of the Ubotica CogniSATTM platform with just one click.

Then, from data in space, these space-borne AI models are leveraged to produce insights. For satellite constellation operators, the new method promises significant CapEx and OpEx savings as well as improved autonomy and decision-making power at the edge with less reliance on base infrastructure.

Space AI applications’ industrialization

Only government-sponsored space programs could realistically produce equipment able to resist the rigors of a rocket launch and the space environment a few decades ago. A new era of commercial space applications is currently being accelerated by entire industries eager to take use of the accessibility of complicated space technology and insights.

Examples include the utilization of reusable rocket launchers and open-source machine learning pipelines. Commercially available off-the-shelf technology, or “COTS,” has been embraced more lately. For a range of applications, such as the in-orbit servicing of satellites and the near real-time detection of forest fires (earth observation), using COTS further lowers the cost of computing in space.

To enable AI inference on data in space, Ubotica is supplying its CogniSAT space AI computing platform on-board satellites. Petabytes of data are daily collected by earth observation firms. The fact that the data is processed and kept locally creates a very significant data difficulty. Customers must develop insights using edge-based data processing analysis in order to get timely application-specific value from all of this data.

The issue of evaluating and managing such large amounts of data must be addressed, as with any other data and AI use cases. In order to tackle this data burden, Ubotica has teamed with IBM to accelerate client deployment of space AI applications and ground-based cloud data processing processes.

This industrialization of artificial intelligence in space necessitates an organizational strategy with fit-for-purpose data storage, governance, and deployment capabilities that are safe, scalable, and simple to use. This collaboration with Ubotica is supported by IBM’s enabling technologies and a commitment to these values.

AI and hybrid clouds will power satellites in the future

In the past, satellites orbited the planet with only basic capabilities to determine when to do things like open the imaging equipment’s aperture or send and receive data based on their position and time in orbit. Almost minimal capacity to discern other situational or contextual factors was used in this.

The European Space Agency and NASA JPL have been receiving space AI capabilities from Ubotica since 2020. CogniSAT-6, the first satellite to employ space AI to autonomously schedule picture tasking and produce insights onboard using image sensor data, was unveiled by Ubotica in 2023.

They are expanding on their significant orbital achievements, which include the Phi-Sat-1 Mission of the European Space Agency and an honorable contribution to the NASA JPL’s ISS Space AI research platform. These technologies will surely have an impact on all facets of the space business by being optimized to function on small satellites.

As a part of this leadership, Ubotica and IBM collaborated to develop a number of services on the IBM Cloud. To enable customer development teams to immediately deploy their AI inference models to Ubotica’s space AI systems aboard LEO satellites is the goal.

A Red Hat OpenShift Kubernetes Service (ROKS) cluster is the initial installation, and it is on this platform that Ubotica will deploy the components for its hybrid cloud AI platform. The first architecture of Ubotica will also incorporate elements from the Open Data Hub reference architecture and is compatible with the watsonx.ai implementation as a whole. Over the course of the upcoming year, Ubotica will also investigate the utilization of additional services from IBM’s Cloud Pak for Data portfolio for integration into their operations.

Ubotica will have the ability to build and modify individual resources based on customer requirements thanks to the IBM Cloud automated deployment system, auto-scaling capabilities, and access to more than 170 platform services.

Utilizing the IBM Ecosystem as a growth platform

Ubotica has access to IBM technology, as well as business and developer know-how and co-creation assistance, through IBM Partner Plus. The IBM New Partner Accelerator has assisted Ubotica in utilizing IBM’s market demand engine, go-to-market materials, and IBM Cloud credits to help enhance its customer space AI model deployment process as a new member of Partner Plus.

In order to target Ubotica’s customized emulator and hardware technologies, the firms are currently working together to establish an enterprise AI model management solution integrated with industry standard source control resources. To eliminate tiresome and potentially error-prone human procedures in the end-to-end delivery of an AI model to an in-orbit asset, automation and observability are offered. The solution will operate with little reconfiguration on multi- and hybrid-cloud deployments.

This strategy is putting Ubotica and IBM in a position to quickly assist shared customers in delivering AI models for use in space applications.

News source: