I've seen a bit of research that relies on imaging methods like FLIR for input. My question adds a couple wrinkles: How could a robot Navigate without imaging? I.e., using only local sensors like a thermocouple or array of thermocouples. Extract navigation info from turbulent thermally heterogeneous media? I'm thinking about some research showing that lobsters and other marine animals may use their antennae to read fine changes in water temp, but I'm at a loss to come up with how they might extract usable info without sitting still to average temperature differences across antennae and estimate the thermal gradient. By all appearances, they don't do this.