Deep reinforcement learning for the olfactory search POMDP: a quantitative benchmark

The olfactory search POMDP (partially observable Markov decision process) is a sequential decision-making problem designed to mimic the task faced by insects searching for a source of odor in turbulence, and its solutions have applications to sniffer robots. As exact solutions are out of reach, the challenge consists in finding the best possible approximate solutions while keeping the computational cost reasonable. We provide a quantitative benchmarking of a solver based on deep reinforcement learning against traditional POMDP approximate solvers. We show that deep reinforcement learning is a competitive alternative to standard methods, in particular to generate lightweight policies suitable for robots.

Automated summary

This article introduces a sequential decision-making problem that simulates the task of insects searching for odor sources in turbulence and explores the application of its solutions to sniffer robots. Due to the infeasibility of exact solutions, the challenge lies in finding the best possible approximate solutions with reasonable computational cost. The researchers provide a quantitative benchmarking of a deep reinforcement learning-based solver against traditional POMDP approximate solvers, demonstrating that deep reinforcement learning is a competitive alternative for generating lightweight policies suitable for robots.