Automated Planet Finder

Lore & Background

The Automated Planet Finder was originally scheduled for first light in 2006, but delays in the construction of major components pushed this back to August 2013, when it was commissioned. Its construction involved international collaboration: the mirror blank was cast in Russia, optically figured in Maryland, USA, and assembled in Arizona, USA; the dome was built in Australia; and the spectrograph was designed and built in California. The telescope examines about 10 stars per night and, over a decade, is expected to study 1,000 nearby stars for planets.

The telescope uses the Ken and Gloria Levy Doppler Spectrometer, which early tests show meets the design goals of high throughput and a radial velocity precision of 1.0 m/s, similar to that of HARPS and HIRES. The design goal is to detect stellar motions as small as one meter per second, comparable to a slow walking speed.

In addition to its exoplanet search, the APF is used to search for optical signals from laser transmissions from hypothetical extraterrestrial civilizations, as part of the Breakthrough Listen project of the Berkeley SETI Research Center.

Reader's Guide

The Automated Planet Finder is notable as a fully robotic telescope dedicated to detecting exoplanets in the five-to-twenty Earth-mass range via radial velocity measurements. Its design precision of 1.0 m/s allows it to detect stellar wobbles as small as a slow walking pace, targeting stars within about 100 light years. The telescope's construction involved international components, reflecting a distributed manufacturing model. Its commissioning in August 2013, after delays from an original 2006 target, marked the start of a decade-long survey of 1,000 nearby stars. The APF's dual use for exoplanet detection and SETI, under the Breakthrough Listen project, extends its scientific scope. The Ken and Gloria Levy Doppler Spectrometer's performance meeting design goals underscores the telescope's capability to contribute to exoplanet science, comparable to instruments like HARPS and HIRES. The project's total cost of $12.37 million highlights its relatively modest budget for a 2.4-meter-class robotic observatory.

Did You Know?

• The telescope is also known as the Rocky Planet Finder.
• Its mirror blank was cast in Russia and optically figured in Maryland, USA.
• It is used for SETI as part of the Breakthrough Listen project.

Defining the Robotic Telescope

A robotic telescope is fundamentally an astronomical instrument and its detector system that conducts observations without any human intervention during the process. The key qualification is that the telescope performs its observations autonomously, even if a human operator must initiate the sequence at the start of a night or terminate it by morning. Some systems incorporate software agents powered by artificial intelligence to assist with tasks like automatic scheduling. It is important to distinguish a robotic telescope from a remote telescope, though a single instrument can occupy both categories simultaneously. By 2004, robotic observations had come to account for an overwhelming share of published scientific findings in several key areas, including asteroid orbit determinations and discoveries, variable star research, supernova light curve measurements and new supernova discoveries, comet orbital tracking, and gravitational microlensing observations. Additionally, every early-phase gamma-ray burst observation in the field was carried out by robotic telescopes, underscoring their critical role in time-sensitive astronomy.

Engineering the Autonomous System

Robotic telescopes are complex assemblies that typically integrate multiple subsystems working in concert. These include mechanisms for telescope pointing, operation of the detector (usually a CCD camera), management of the dome or enclosure, control of the focuser, and weather-condition detection, among other capabilities. A master control system, almost invariably a software component, oversees these varying subsystems. The operating philosophy divides into two categories: open loop and closed loop. In an open loop configuration, the telescope points itself and gathers data without ever inspecting its own results to verify proper operation. Such a system is sometimes described as operating on faith, because if something malfunctions, the control system has no mechanism to detect the error or compensate. A closed loop system, by contrast, evaluates its own operations through redundant inputs to catch errors. Common examples include position encoders mounted on the telescope's axes of motion, or the ability to assess the system's images to confirm it was aimed at the correct field of view during exposure. Most robotic telescopes tend to be small instruments, while large observatory equipment may be highly automated but rarely operates without human attendants.

From Pioneering Experiments to Professional Networks

The earliest robotic telescopes emerged after electromechanical computer interfaces became standard at observatories. These initial examples were costly, limited in capability, and relied on numerous unique hardware and software subsystems, which hampered progress in the field. The landscape shifted by the early 1980s when inexpensive computers became available, enabling several viable robotic projects to be conceived and built. A landmark 1985 book by Mark Trueblood and Russell M. Genet, Microcomputer Control of Telescopes, identified many subtle reasons why telescopes could not be reliably pointed using basic astronomical calculations alone. The University of Iowa has led professional robotic development since the late 1980s, progressing from the Automated Telescope Facility on a physics building roof in the early 1990s, to the Iowa Robotic Observatory at Winer Observatory in 1997, to the 0.37-meter Rigel Telescope completed in May 2002. Today, large networks like RoboNet, operated by a UK university consortium, and the LINEAR project represent the professional frontier, with competitors like Lowell Observatory and the Catalina Sky Survey pursuing varying degrees of automation.

Software Challenges and the ROTSE Legacy

A persistent challenge in professional robotic telescope development has been software sustainability. By 2004, many systems were characterized by a lack of design creativity and heavy reliance on closed-source, proprietary software unique to each telescope. At universities, this problem was acute: graduate students who wrote the code would move on to new positions, leaving their institutions without the knowledge to maintain the system. Large consortia and government-funded laboratories tended to avoid this attrition. Competition for research funding between institutions often drove the need for proprietary solutions. Meanwhile, the ROTSE program demonstrated the transformative potential of fully autonomous operation. ROTSE-I began in manual mode in 1997, achieved full robotic operation in late March 1998, and became the first fully autonomous closed-loop robotic telescope. It captured the first prompt optical burst from a gamma-ray burst, GRB 990123. The follow-on ROTSE-III, with four half-meter telescopes operational from 2003, was used for GRB follow-up and supernova searches, and it was through ROTSE-III observations that the first superluminous supernovae were discovered.