The Inevitable Cost of Precision in a Noisy Environment
Prof. Dr. Udo Seifert (Universität Stuttgart)

All processes in cell and molecular biology, like transport by molecular motors and replication machinery, are subject to thermal noise. Still, life relies on the fact that the result of such processes comes with a small enough uncertainty, i.e., large enough precision. The same holds for (wet) micro- and nano-robotics. While absolute precision is impossible in an environment of finite temperature, an obvious question is whether or not there is a fundamental trade-off between precision and the (free energy) cost of generating or running such processes.

After recalling the principles of stochastic thermodynamics, I will introduce the recently discovered thermodynamic uncertainty relation that provides a universal lower bound on the precision any process in steady-state conditions can achieve for a given energy budget. A variant of this relation allows us to extract from experimental data a model-free upper bound on the efficiency of molecular motors. Likewise, for steady-state heat engines, this relation shows that Carnot efficiency can be reached at finite power, in principle, but only at the cost of diverging power fluctuations. I will close with recent insights into the minimal requirements for generating coherent oscillations in (biochemical) networks.