Physical mobility analytics have gained attention lately. As people become more equipped with ubiquitous wireless-communication-enabled mobile appliances, they tend to leave signatures of their presence wherever they go. One particular example is Wi-Fi enabled devices which continuously send packets (called “probe requests”) to access points around it even if no connection is established between them. Aggregating a list of such probe requests over a number of geographically distributed monitoring nodes gives rise to a rich set of physical mobility analytics such as visitor density in rush hours and most frequently taken routes. However, privacy of individual users is a grave concern. To address this concern we propose to implement physical mobility analytics using a collection of privacy-preserving primitives of set operations. The sets are the MAC addresses of the devices observed by one monitoring node. There is at least one set per monitoring node. An monitoring node may have more than one set if the MAC addresses are split according to the time of reception. The primitives we propose are the t-incidences of these sets. We present an ε-differentially pan-private algorithm to estimate the t-incidence of n sets, up to multiplicative error O(α), given three (ε/3)-differentially pan-private Bloom filters for each of those sets.