Gas Source Tracing With a Mobile Robot Using an Adapted Moth - - PowerPoint PPT Presentation

▶

Aug 30, 2023 113 likes •350 views

Gas Source Tracing With a Mobile Robot Using an Adapted Moth Strategy Achim Lilienthal, Denis Reimann, Andreas Zell University of Tbingen, WSI Contents 1) Gas Source Tracing Strategy of Bombyx mori 2) Gas-Sensitive Robot 3) Adapted

SLIDE 1

Gas Source Tracing With a Mobile Robot Using an Adapted Moth Strategy

Achim Lilienthal, Denis Reimann, Andreas Zell University of Tübingen, WSI

SLIDE 2

2 Achim Lilienthal (Universität Tübingen, WSI-RA, Prof. Andreas Zell)

1) Gas Source Tracing Strategy of Bombyx mori 2) Gas-Sensitive Robot 3) Adapted Gas Source Tracing Strategy 4) Experimental Setup 5) Results 6) Summary & Outlook

SLIDE 3

3 Achim Lilienthal (Universität Tübingen, WSI-RA, Prof. Andreas Zell)

Gas Source Tracing of Bombyx mori

1 Fixed Movement Pattern

triggered by increased pheromone concentrations surge towards upwind direction

Kanzaki 1998

SLIDE 4

4 Achim Lilienthal (Universität Tübingen, WSI-RA, Prof. Andreas Zell)

Gas Source Tracing of Bombyx mori

1 Fixed Movement Pattern

implement a local search for the next patch

Kanzaki 1998

SLIDE 5

5 Achim Lilienthal (Universität Tübingen, WSI-RA, Prof. Andreas Zell)

Gas Source Tracing of Bombyx mori

1

“Remarkably, the movement tracks that many species produce when tracking odour plumes have a very similar side-to-side zigzag shape whether walking, swimming, or flying.” [Mark Willis]

Kanzaki 1998

SLIDE 6

6 Achim Lilienthal (Universität Tübingen, WSI-RA, Prof. Andreas Zell)

Experimental Setup - RoboMoth

2 “Arthur” (ATRV-Jr.)

footprint = 80 x 65 cm height = 55 cm

Sensors Used

dometry

laser range scanner

SLIDE 7

7 Achim Lilienthal (Universität Tübingen, WSI-RA, Prof. Andreas Zell)

Experimental Setup - RoboMoth

2 Gas Sensitive System

based on VOC-Vario using two TGS 2620

SLIDE 8

8 Achim Lilienthal (Universität Tübingen, WSI-RA, Prof. Andreas Zell)

Adapted Gas Source Tracing Strategy

3 Wind Direction not Known Exploit Relative Sensor Response

SLIDE 9

9 Achim Lilienthal (Universität Tübingen, WSI-RA, Prof. Andreas Zell)

Adapted Gas Source Tracing Strategy

3 Adapt Threshold (Compensate Slow Response/Recovery)

SLIDE 10

10 Achim Lilienthal (Universität Tübingen, WSI-RA, Prof. Andreas Zell)

3 How to Calculate x ?

differences between individual sensors different environmental conditions increasing base level of concentration

xt = Rt − Rmin,t Rmax,t − Rmin,t

Data Preprocessing

SLIDE 11

11 Achim Lilienthal (Universität Tübingen, WSI-RA, Prof. Andreas Zell)

3 Update Minimum and Maximum Dynamically

trimming by 1% every 30 s

Data Preprocessing

SLIDE 12

12 Achim Lilienthal (Universität Tübingen, WSI-RA, Prof. Andreas Zell)

Experimental Setup - Environment

4 Office Environment

natural indoor environment (to some extent)

SLIDE 13

13 Achim Lilienthal (Universität Tübingen, WSI-RA, Prof. Andreas Zell)

Results

5 Active Source Present

SLIDE 14

14 Achim Lilienthal (Universität Tübingen, WSI-RA, Prof. Andreas Zell)

Results

5 Active Source Present

SLIDE 15

15 Achim Lilienthal (Universität Tübingen, WSI-RA, Prof. Andreas Zell)

Results

5 No Active Source Present

SLIDE 16

16 Achim Lilienthal (Universität Tübingen, WSI-RA, Prof. Andreas Zell)

Results

5 Statistics (595 min)

average distance median distance active source, all

(225.6 ± 179.6) cm 266.2 cm

inactive source

(317.7 ± 235.2) cm 336.8 cm

SLIDE 17

17 Achim Lilienthal (Universität Tübingen, WSI-RA, Prof. Andreas Zell)

Results

5 Statistics

after 1. trigger

SLIDE 18

18 Achim Lilienthal (Universität Tübingen, WSI-RA, Prof. Andreas Zell)

Results

5 Statistics (595 min)

active source, after first trigger average distance median distance active source, all

(225.6 ± 179.6) cm 266.2 cm

inactive source

(195.5 ± 119.8) cm (317.7 ± 235.2) cm 170.9 cm 336.8 cm

SLIDE 19

19 Achim Lilienthal (Universität Tübingen, WSI-RA, Prof. Andreas Zell)

Results

5 Statistics (595 min)

active source, after first trigger average distance median distance active source, all

(225.6 ± 179.6) cm pH0 = 0.433

inactive source

(195.5 ± 119.8) cm (317.7 ± 235.2) cm pH0 = 0.173

SLIDE 20

20 Achim Lilienthal (Universität Tübingen, WSI-RA, Prof. Andreas Zell)

Results

5 Statistics (595 min)

active source, after first trigger average distance median distance active source, all

pH0 = 0.167 266.2 cm

inactive source

pH0 = 0.167 170.9 cm 336.8 cm

SLIDE 21

21 Achim Lilienthal (Universität Tübingen, WSI-RA, Prof. Andreas Zell)

Summary & Outlook

6 Adapted Biomimetic Strategy (Bombyx mori)

local search (fixed motion pattern) triggering mechanism

Suitability for Indoor Gas Source Tracing

good evidence

More Experiments needed Larger Environment

SLIDE 22