Atomic clocks have been at the leading edge of accuracy and precision since their inception in the 1950s. However, typically the most capable of these clocks have been confined to laboratories despite the fact that there are compelling reasons to apply them in the field and/or while in motion. These applications include synchronization of distributed critical infrastructure (e.g. data servers, communications, electricity grids), scientific applications (e.g. radio-astronomy) and to mitigate the effects of interruption to Global Navigation Satellite Systems. Over the last 20 years, there has been a breakthrough in the performance of atomic clocks by transitioning from an atomic reference based on a microwave transition to an optical frequency transition. The $10^5$-fold increase in reference frequency confers the potential for significantly higher performance. However, this performance increase has come at the cost of size, complexity and fragility which has continued the confinement of these clocks to the laboratory. Here we report on a recent international collaboration where three emerging optical clocks, each operating on different principles, were trialed at sea. These clocks incorporate optical frequency combs so that their stable frequency outputs can be used directly in electronic apparatus and were also automated so that they do not require expert supervision. We present the frequency stability and reliability of these three clocks over multiple weeks of unsupervised naval trials, both in harbour and on the ocean. The performance of all three devices was orders of magnitude superior to existing best-in-class commercial solutions over short and medium timescales. This demonstrates that optical clocks are ready to deliver advantages for real-world applications.
