Autonomous astrophoto scheduling with Stella AI

Advancements in astronomy and astrophotography have increasingly revolved around automation, integration, and artificial intelligence. The StellarMate platform exemplifies this trend through its latest innovation, Stella AI, a virtual assistant designed to streamline and fully automate astrophotography sessions. In a demonstration presented by Jassem Mutlaq, the creator of StellarMate, viewers are shown how to use Stella AI within the StellarMate mobile application to execute a complete imaging session. The demonstration revolves around scheduling the Christmas Tree Cluster for a fully autonomous session, and it highlights the app’s seamless user interface, Stella AI’s intuitive command parsing, and the robust integration with the Ekos platform.

Autonomous astrophoto scheduling: Use Stella AI to schedule your fully autonomous sessions in the StellarMate app

The walkthrough begins with the fundamental requirement: the StellarMate app must be connected to the internet. Stella AI functions through cloud interaction, requiring online access to communicate with the backend services that parse user commands and generate executable schedules. Jassem emphasizes that the version of Stella AI in the StellarMate app is identical in capability to the one accessible via the EkosLive web portal. This interoperability ensures that users receive a consistent experience regardless of the platform they choose to operate from. Whether using a mobile device or a desktop browser, Stella AI understands the same prompts and follows the same decision-making logic.

Once inside the StellarMate app, users navigate to the Stella tab. Jassem identifies his imaging target for the night, the Christmas Tree Cluster. While the app has a traditional method for scheduling sessions that involves navigating to the Targets tab, selecting an object, pressing the Schedule button, and manually selecting a sequence file, this method can be cumbersome, especially for beginners or those unfamiliar with the various file parameters required. By contrast, Stella AI condenses the entire process into a single spoken or typed instruction.

Stella AI supports two primary operational modes, depending on the user's intentions. If a session is already active and the goal is simply to collect more images, the user begins the prompt with the word "capture." For example, they might say, "capture 20 dark frames at minus 20 degrees Celsius." However, if the objective is to orchestrate a complete automated session from initialization to shutdown, the keyword to use is "schedule." This distinction allows Stella AI to understand the broader context of the user’s goal and respond with the appropriate type of plan.

Commanding Stella AI and verifying session parameters

In the demonstration, Jassem proceeds with the scheduling method. He activates the microphone feature within the app and speaks a single command: "Schedule the Christmas Tree Cluster and capture 20 images, each 300 seconds long, in hydrogen alpha, sulfur, and oxygen. Make sure the binning is 2x2." After speaking the prompt, he presses enter to confirm. Within moments, Stella AI parses the command and generates a plain-language summary of the intended task. This includes the target object, the exposure time, the number of frames, the specified filters, and the binning mode. The system then prompts the user to either tap or say “execute” to proceed with initiating the session.

Jassem takes a moment to emphasize the presence of two files that Stella AI generates: the schedule file and the sequence file. These files are fundamental to the functioning of the automation system. The schedule file governs the overall session behavior, including which object to observe, when to begin, and how to handle startup and shutdown procedures. The sequence file, meanwhile, contains detailed information about each image capture task, exposure lengths, binning settings, filter choices, and other parameters critical to data collection. These files are dispatched to Ekos, the software framework that handles all observatory automation tasks.

Before proceeding with execution, Jassem recommends inspecting the files to ensure accuracy. Upon reviewing the generated schedule, he verifies that the Christmas Tree Cluster has been correctly identified as the target. He also checks that the startup and shutdown procedures for the observatory are correctly configured. These procedures might include warming up or cooling down specific equipment, opening or closing the observatory roof or dust cover, or homing the telescope mount before slewing to the target. Such preparatory steps are essential for ensuring equipment safety and session accuracy.

Moving to the sequence file, Jassem confirms that the file reflects the desired 300-second exposure time, 2x2 binning, and use of hydrogen alpha, sulfur, and oxygen filters. These filters are commonly used in narrowband astrophotography and are ideal for capturing the fine details of emission nebulae like those found in the Christmas Tree Cluster. Satisfied that everything is correctly configured, Jassem presses the execute button.

At this point, Ekos takes over, beginning with the activation of the equipment profile. This triggers a cascade of tasks: connecting to the mount, camera, filter wheel, focuser, and other instruments; unparking the telescope from its rest position; and initiating startup sequences. Within the app’s Observatory View, Jassem observes the real-time status as the mount begins to unpark. He notes that the dust cover is also being automatically removed, a critical step before imaging can begin.

Telescope alignment, guiding, and image capture execution

As the automation progresses, Ekos aligns the telescope with the night sky, using plate solving or other alignment procedures to precisely locate the Christmas Tree Cluster. Once aligned, the mount slews to the target’s coordinates. Focusing begins, often involving either an autofocus routine or a pre-saved focus position for the specific filter in use. Stella AI does not perform these technical tasks itself; instead, it serves as the orchestrator, creating the schedule and instructing Ekos to carry out the physical operations.

Guiding is another key step in the process. Once the mount is aligned and tracking the object, Ekos engages the autoguider to ensure that the telescope remains precisely locked on the target throughout each long exposure. This is essential for avoiding star trails or image distortion due to minor tracking errors. After guiding is successfully established, the camera begins its imaging sequence.

Jassem does not go into exhaustive detail about the image capture phase, but the implication is clear: from the moment Stella AI receives its command and generates the files, every other aspect of the session proceeds automatically. The imaging session unfolds based on the parameters provided, with no further intervention from the user required. Each 300-second exposure is taken with precision, and the filter wheel automatically rotates between hydrogen alpha, sulfur, and oxygen as specified.

Finalizing the session and the vision behind Stella AI

At the end of the session, the schedule file includes shutdown procedures that bring the system safely back to a resting state. This may include parking the mount, closing the dust cover, warming the camera sensor, and disconnecting equipment to prevent wear or damage. The entire session, from start to finish, is self-contained within a single schedule, making it possible for astrophotographers to set up and walk away, knowing that Stella AI and Ekos will manage the operation responsibly.

The demo concludes with Jassem reminding viewers that this was a quick demonstration of how to use Stella AI for full autonomous scheduling with StellarMate. He signs off with a wish for clear skies, the universal blessing among amateur and professional astronomers alike.

Ultimately, this demonstration encapsulates the broader vision behind StellarMate and its AI assistant. Automation in astronomy is not about removing the joy of discovery or the satisfaction of crafting an image by hand—it is about eliminating repetitive tasks, reducing human error, and enabling astrophotographers to focus on creativity, data analysis, and scientific exploration. With tools like Stella AI, users are empowered to plan and execute complex imaging sessions with just their voice, turning what once required hours of planning into a streamlined experience. As artificial intelligence continues to evolve within astronomy, the barriers to entry for astrophotography are being lowered, inviting more people to explore and document the wonders of the night sky.