Telescope Control and Robotics


For many decades, observational astronomy meant a professional astrophysicist traveling to a proper telescope and conduction research in situ. Observing slots were typically allotted in shares of a few days. Some science questions are hampered by such a procedure. Long-term campaigns spanning month or sometimes years do not fit into this schema, even if only a few minutes of observing time is needed in any given night. Targets of opportunity are also hard to tackle. It requires the lucky turn of the card to be on spot, if something unforeseen happens. Survey observing mode, where a staff of professional observers reside at the telescope was introduced at all major facilities to mitigate these problems. Still, proposals are normally evaluated on semester basis, making year-long campaigns difficult - or at the courtesy of the TAC. Really fast follow up required, e.g., on gamma-ray bursts, remain impossible even in service mode. Overcoming such shortcomings is the domain of autonomous observing, where the AIP has a long-standing experience in.

The section Telescope Control and Robotics focuses on the development of autonomous telescopes and robotic observatories through innovative software solutions. Additionally, we provides special-purpose hardware for telescope control beyond the standard equipment.

How do you optimally fill a Knapsack?

Such seemingly simple questions may have complex answers. As long as the knapsack is small, a simple trial-and-error approach leads to the true answer, but what happens if your knapsack can hold a hundred items and you may choose among thousand? Trial and error leaves you with 1000 choose 100, nCk(1000,100) or more than (far more!) 10^100 possibilities. If you can do one trial in 1ns, the trial-and-error time exceeds the age of the universe. But there are ways to mitigate such problems...(read more)

Make observing more efficient by getting rid of the human factor

Astronomical observations can be a tedious task, especially if repetitive tasks are to be conducted over the course of an entire night. Fatigue may lead to errors, valuable observing time might be lost. Thus, autonomous observing tries to pull the human factor out of the loop; a robotic observatory has also control over the building and does not require human intervention at all. Though in cybernetics, the term robotic requires the system to show self-learning aspects, astronomers tend to name anything capable of automated observing a robotic telescope. But how can you pull the human out of the loop? Well, start with careful considering how an observation can be split into individual subtasks...(read more)

Automated data handling with data reduction pipelines

Robotic observatories are much more efficient in producing scientific valuable data than any human can ever be. Additionally, because calibration and on-object observing strategies occur template-based, the inflow of calibration data and scientific data is steady, making data reduction a task that can easily be automated - nobody can 'forget' to also take calibration data in daytime after an exhausting observing run during nighttime. Though the institute's strategy is to anchor the responsibilities for data reduction within the appropriate section holding the expertise, it is well recognized that data reduction done, e.g., in optical wavelength, where data is recorded on CCD's have some commonalities. Building blocks of data reduction can be interwoven, at the same time opening possibilities to the scientists to choose from different algorithms to conduct specific tasks. ESO offers their Reflex environment to achieve a similar goal...(read more)

Core projects:


When STELLA (short for STELLar Activity) was inaugurated in 2006, it was the first >1m class robotic telescope feeding a high-resolution spectrograph in the world. The combination with a second 1.2m telescope used for imaging is still rather unique and makes it apt for many different types of observing programs. Constant improvements in the acquisition and guiding, along with slew-time minimizing scheduling optimizations, allowed us to reach a shutter-open time of slightly more than 90% in the spectrograph, and more than 55% in the imaging telescope. The system uptime constantly surpasses 98%, and since 2018, more than 100 users have been granted time shares...(read more)


BMK10k (Ballistische Messkammer) is a so-called Astro-Topar system with 30cm diameter and a flat field of view of 20° diameter. It is equipped with a 10kx10k detector, a spare CCD from the PEPSI project, which still sees a 7.25°x7.25° field. In 2019 it has been shipped to Cerro Aramzones (neighboring the E-ELT!) and successfully saw first light there in August the same year. The BMK10k will mainly support ESA’s PLATO mission, a planetary transit search experiment covering an utterly large 2230 square degree field-of-view. BMK10k will provide a three-year long survey with photometry of all stars in the South field down to 17-18th mag, but at a resolution 6 times better than PLATO...(read more)

Keeping the LBT on track

This section was also responsible for building six Acquisition, Guiding & Wavefront-sensing (AGW) units for the LBT, on-axis parts were built by Arcetri Observatory INAF. These units provide real-time information on the dynamic state of the atmosphere, in turn allowing the Adaptive Optics system to compensate for the telescope’s own optical defects...(read more)

A common operating software for robotic telescopes

The software operating all AIP’s robotic telescopes, named after its first use case STELLA Control System (SCS) is entirely written in Java and thus easily portable to other systems. The general software layout has always been highly modular as a consequence of the first use case -- operating a spectroscopic telescope and an imaging telescope together. We claim that it can be adapted to any robotic telescope with relative ease...(read more)

Other projects include:

  • For PEPSI, together with the high-resolution spectroscopy group we built SDI, the solar disc integration telescopes and its successor SDI-POL. Also the PFUs, the permanent fiber units are a close collabration with the  high-resolution spectroscopy group.
  • For the ARGOS Laser system at the LBT, we lead the sub-project on the tip-tilt sensors.
  • Automating the VATT for feeding light of the Vatican Advanced Technology Telescope via an optical fiber into PEPSI.
  • RoboTel: A smaller STELLA clone at the Campus Babelsberg for instrument testing and hopefully soon open to the public.
  • APT7 (Amadeus): An automated photoelectric telescope operated since 1995 at Fairborn observatory, Az, USA. Now retired.
  • ICE-T: Originally intended as a replacement for the canceled Eddington ESA-mission for Dome-C, Antarctica. Discontinued due to unclear access situation on Dome-C.
Last update: 30. August 2021