cs.ROOct 25, 2026

The German Tank Problem & Navigating by Scent

Authors: Kordel Kade France

Organizations: Scentience, Inc.

Read PDF
Ask AI
Cite

Copy citation

DOINot available for this paper

Abstract

A robot searching for the source of an odor must eventually answer a deceptively hard question: have I arrived? Near the source, concentration gradients flatten, turbulence makes the signal intermittent, and a small robot often lacks the positional accuracy to map the plume. Here, we describe a lightweight remedy inspired by a World War II statistics conundrum: the German tank problem. To approximate enemy armament production, Allied analysts developed an estimator based on serial numbers from captured tanks. Analogously, we use the same procedure over sampled chemical sensor readings from a robot to create estimates for the peak concentration of an odor plume, and therefore the odor's source. Treating each sensor reading as an observation, our unbiased estimator needs only a running maximum and a sample counter. An exact confidence bound turns both numbers into a principled rule that the robot or UAV can use to determine whether to keep searching for the plume source or land. We derive the estimator and its bound, demonstrate its sample efficiency, and examine its robustness even when its assumptions fail. Finally, we summarize the estimator's role as the termination criterion for olfactory navigation.

Figures & tables

Figure 1: The German tank problem uses the same estimator as a robot determining when it has reached the source of an odor plume.
VariableGerman tank problemOdor source declaration
Populationtanks producedchemical sensor readings
Label on each unitserial numberchemical concentration
Unknown upper endtotal number of produced tanks, NNmax concentration at the plume source, C⋆C^{\star}
Observationskk captured tankskk sensor readings
Samplingwithout replacementwith replacement (readings of a given strength recur)
Sufficient summarykk and the largest serial number, mmkk and the running max concentration, mkm_{k}
Table 1: The German tank problem and olfactory navigation share a common thesis: estimate the upper end of a range from the largest value sampled so far.
Appendix figures & tables2 assets

Supplementary material from the paper’s appendix.

Appendix

Figure S1: How fast the evidence accumulates. Left: the bound as a multiple of the running maximum, Uk/mk=α−1/kU_{k}/m_{k}=\alpha^{-1/k} , and the point-estimate factor (k+1)/k(k+1)/k . At α=0.025\alpha=0.025 , as in [ 8 ] , the bound falls below 1+ε=1.251+\varepsilon=1.25 at k⋆=17k^{\star}=17 . Right: the same factors as a firmware lookup table ( 201/k20^{1/k} and 401/k40^{1/k} are α−1/k\alpha^{-1/k} at α=0.05\alpha=0.05 and 0.0250.025 ), with the relative standard deviation of C^k\widehat{C}_{k} .
ModalitySensorPolicyMean (s)Std. dev. (s)Best (s)
OlfactionMOXOIO98.3814.8471.59
OlfactionECOIO112.2216.7097.65
OlfactionMOXE. SARSA( λ\lambda )103.9914.8982.73
OlfactionECE. SARSA( λ\lambda )121.8516.7797.10
Olfaction + VisionMOXOIO94.0314.6866.12
Olfaction + VisionECOIO107.7117.0780.50
Appendix
Table S1: Navigation times for a UAV navigating a course to localize the source of a chemical compound using olfaction only and olfaction-vision hybrid sensing, as reported in [ 8 ] . Each trial used the unbiased estimator presented in this paper, and was critical to achieving fast and reliable navigation times. Every trial ended with the UAV landing within 1 m of the source. OIO: olfactory inertial odometry; E. SARSA( λ\lambda ): Expected SARSA( λ\lambda ).
More in this topic

Robot Navigation

Robot navigation studies how robots perceive, localize, plan, and move safely toward goals in static or changing environments.

See all 554 papers

Explore similar work

CardsList

  1. SmellDiffusion: Diffusion-Based Quadruped Navigation with Olfactory Scene Graphs

    Sep 17, 2026Faith Ogunwoye, Iana Zhura, Hajira Amjad +5Trajectory GenerationDiffusion-Based Planning

    Skolkovo Institute of Science and Technology, Moscow, Russia

  2. Scensory: Real-Time Robotic Olfactory Perception for Joint Identification and Source Localization

    Sep 11, 2025Yanbaihui Liu, Erica Babusci, Claudia K. Gunsch +1Robot NavigationRobot Localization

    Department of Mechanical Engineering and Materials Science, Duke University, Durham, NC, USA. · Department of Civil and Environmental Engineering, Duke University, Durham, NC, USA. · Department of Electrical and Computer Engineering,Duke University, Durham, NC, USA. +1

  3. Flow structure decides whether a robot searching for odour should follow the wind

    Oct 14, 2026Junlong Shen, Shunsuke ShigakiRobot Navigation

    Department of Electrical and Computer Engineering, University of Alberta, Edmonton, Canada. · Principles of Informatics Research Division, National Institute of Informatics, Tokyo, Japan.