In spite of widespread opinion it is possible to construct a spectral measure covariant with respect to the time translations. This spectral mesure corresponds to the time operator. However, its physical interpretation is not clear.
The criticism formulated by Mielnik and Torres-Vega in their paper, does not apply to my derivation of the time of ar-rival in Quantum Mechanics but to our present understanding of the quantum measurement theory. In paper [1] I have expressed the following opinion: Although, there is no room to modify the mathematically unique definition of the probability QI×D(ψ) within the standard mathemati-cal framework of Quantum Mechanics (spectral measures, self-adjoint operators etc.), the problem is still open from the physical point of view. Indeed, the probabilistic interpretation of the quantum mechan-ics is based on “short ” measurements of the type “What is a proba-bility that the particle will be found within the three-dimensional vol-ume V precisely at a given time instant t? ” How to relate them with “long ” measurements, corresponding to time intervals rather than to time instants t is by no means obvious. Bogdan Mielnik and Gabino Torres-Vega propose a deep analy-sis of conceptual problems which arise when one starts to inter-
We study the positive-operator-valued measures (POVMs) on the projective real line covariant with respect to the projective group. We interpret the projective line as a compactified time axis and we assume that the energy is a positive operator. This formalism may describe a time-of-arrival observable for a free particle covariant with respect to linear canonical transformations. The problem is similar to the more complicated and physically more relevant problem of finding the POVMs on the compactified Minkowski space-time covariant with respect to the conformal group.