The deployment of the U.S. Navy’s first nuclear-powered Polaris ballistic-missile submarine, the USS George Washington (SSBN-598), was a true revolution in naval affairs thanks to her ability to launch missiles while submerged. But several groundbreaking technologies had to be developed for the Polaris system to come to fruition. Perhaps the most important of these was the highly accurate submarine inertial-navigation system (SINS) that could keep track of the submarine’s position without requiring the boat to obtain surface fixes. SINS permitted Polaris submarines to remain underwater for extended periods while providing the missiles with targeting data accurate enough to achieve the one-nautical-mile circular-error-probability precision for which they were designed.
As the Navy moved into the guided-missile era in the late 1940s, it needed a system capable of calculating the explicit geometric relationships among ship coordinates, target coordinates, and horizontal coordinates. This was the objective of Project Mast (short for “marine stabilization”), established on 3 September 1948 when a study contract was issued to the Massachusetts Institute of Technology (MIT) Instrumentation Laboratory. Preliminary tests of the Project Mast hardware led Charles Stark Draper, director of the Instrumentation Laboratory (which was renamed in Draper’s honor in 1970), to conclude it was possible to build an inertial-navigation system for use on board naval vessels. This idea was presented to the Office of Naval Research, which entered into a contract with MIT in June 1950 to study such a system’s feasibility for submarines.
The Instrumentation Laboratory’s August 1951 report reiterated the previous conclusion that “a servo-driven platform on which were mounted two accelerometers (or pendulums) and two gyroscopes, could be used to indicate the vertical, and, with the addition of a third gyroscope, to indicate heading.” This combination placed on board a submarine would show the direction of the Earth’s gravity field and indicate the projection of the planet’s polar axis in the horizontal plane.
The Navy contracted with the lab to develop a prototype. On 6 November 1951, the Inspector of Naval Material, based in Cambridge, Massachusetts, forwarded Bureau of Ships (BuShips) specifications for equipment that would permit accurate navigation of submarines over a period of ten hours without requiring the vessel to disclose its presence to hostile forces.
Engineer Forrest Houston was placed in charge of the project. Houston, a U.S. Naval Academy graduate who had served in the Pacific during World War II as a gunnery officer, was the ideal candidate to manage the SINS project. He had received his master’s degree in electrical engineering from MIT in 1948 while still in the Navy. His thesis, “Stabilization and Tracking Control of a Proposed On-mount Antiaircraft Fire Control System,” covered background material that would be useful for Project Mast. He was hired by the Instrumentation Laboratory in 1950 and assigned to the initial study of submarine navigation.
The first SINS prototype was completed by the end of 1953. It was tested on shore in January 1954, in a van driven from Boston to Newbury, and a number of runs were made over roads across the eastern United States. The first shipboard tests were conducted on board a fleet oiler, the USS Canisteo (AO-99), during a 15-day trip from Norfolk, Virginia, to Long Island, New York, and back. In February 1955, the system was installed on the cargo ship USS Alcor (AK-259) and tested on a roundtrip voyage from Norfolk to Naples, Italy. Additional test runs were carried out in the Mediterranean and Caribbean. Draper reported afterward that “the results, in terms of latitude errors and longitude errors, were erratic and tended to have mean values in the vicinity of 5 nm.” Although good performance was obtained for a period of as long as 108 hours, operation was intermittent and positions tended to drift.
As a result of these trials, BuShips revised the design requirements for SINS and requested bids from commercial firms for an experimental model for submarines. Two firms responded and were awarded contracts: One went to the Autonetics Division of North American Aviation and the other to the Marine Division of Sperry. The Sperry system, delivered by September 1956, was based on the MIT design, in which the gyroscopes remain in fixed orientation in inertial space with respect to the stars. This created unforeseen problems for Sperry, whose capabilities in precision manufacturing were not as good as those of MIT’s Instrumentation Laboratory. As Graham Spinardi wrote in his history of the Fleet Ballistic Missile Program:
This led to a system that rapidly lost accuracy. As the earth rotated and the submarine’s position changed, the gyroscopes were subjected to a varying gravity field. The slightest mass imbalance of their rotors would lead to significant errors. But achieving perfect or near perfect mass balance was an exceedingly difficult task, especially as one moved outside the laboratory.
The North American Autonetics system, the N6A Autonavigator, originally had been developed for the North American SM-64 Navaho, a supersonic intercontinental cruise missile. Unlike the Instrumentation Laboratory’s design, it was a local-level system kept horizontal at all times so the gyros were not subject to changes in direction of gravity. It also incorporated an innovative approach to the drift problem developed by John Slater, a key figure in inertial-instrument design at Autonetics. Each of the three axes had two G2K gyros that could be reversed, averaging out the drift of the gyros. They were operated in alternating sets—spinning them in one direction for a time, and then passing control to the second set, permitting the original pair time to spin down and spin up in the opposite direction. This reversing spin-up, spin-down cycle was called a NAVAN cycle.
To compare the performance of Sperry’s SINS Mark 1 with a version of the North American N6A-1, the Navy’s Special Projects Office conducted sea trials on the USS Compass Island (EAG-153). The Compass Island had been converted and classified as a navigational research test vessel under the Polaris missile system budget. The ship operated along the Eastern Seaboard testing equipment and training personnel until 13 March 1958, when she sailed from New York for experiments in the Mediterranean, returning to New York on 17 April. The Autonetics system performed much better than the Sperry system and was installed in the USS Nautilus (SSN-571), where it played an essential role in guiding the boat under the ice during the Nautilus’s record-breaking voyage to the North Pole in August 1958.
Despite the performance advantage exhibited by the Autonetics system, the Navy kept both systems in development. Each was modified and improved with the help of the MIT Instrumentation Laboratory, which continued to provide technical support to both companies under a separate Navy contract. It is likely that Sperry, which had been selected to integrate all the navigation equipment for Polaris submarines (three SINS, a stabilized periscope for celestial navigation, a radio navigation system, and two navigation data assimilation computers [NAVDACs]), lobbied hard for the continued development of its production version of SINS, designated the Mark 3.
The Navy initially intended to install Sperry’s SINS Mark 3 in the first five George Washington–class Polaris submarines, with the Autonetics version, SINS Mark 2, allocated to the newer Ethan Allen class. When Autonetics was able to produce its SINS system faster than Sperry, the Navy reversed the schedule. The Sperry systems all eventually were replaced with the Autonetics Mark 2 (and its various descendants), which became standard on all the Navy’s fleet ballistic missile submarines. A total of 56 Mark 2 Mod 0 SINS were produced, with three redundant systems installed in each submarine.