Application of learning from demonstration to a mining tunnel inspection robot: The inspection of muck piles and other entities in mines is of keen interest to mining engineers and project managers. This paper presents initial work on developing an automated robot system capable of such inspections. The robot system uses a robot training technique called Learning from Demonstration (LfD), where after human demonstrations a robot can finish a pre-defined task. The paper presents research on decomposing a task into modules that increases robot adaptability in learning an inspection task. The paper shows how modularization is appled to a robot inspecting a muck pile using a Gaussian Mixture Model (GMM) togother with Gaussian Mixture Regression (GMR). Results of simulations and real robot perfromance are presented.

Feng Lu Ge; Wayne Moore; Michael Antolovich

Application of learning from demonstration to a mining tunnel inspection robot: The inspection of muck piles and other entities in mines is of keen interest to mining engineers and project managers. This paper presents initial work on developing an automated robot system capable of such inspections. The robot system uses a robot training technique called Learning from Demonstration (LfD), where after human demonstrations a robot can finish a pre-defined task. The paper presents research on decomposing a task into modules that increases robot adaptability in learning an inspection task. The paper shows how modularization is appled to a robot inspecting a muck pile using a Gaussian Mixture Model (GMM) togother with Gaussian Mixture Regression (GMR). Results of simulations and real robot perfromance are presented.

Feng Lu Ge, Wayne Moore, Michael Antolovich

Research output: Book chapter/Published conference paper › Conference paper › peer-review

Original language	English
Title of host publication	2011 CRCMining: Australian Mining Technology Conference
Place of Publication	Kenmore East, Qld
Publisher	CRC Mining
Pages	28-35
Number of pages	8
Publication status	Published - 2011
Event	2011 CRCMining Australian Mining Technology Conference - Australis Noosa Lakes Resort, Tewantin, Noosa, Australia Duration: 12 Sept 2011 → 14 Sept 2011 https://web.archive.org/web/20110723024407/http://www.crcmining.com.au/dynamic_page.php?page_reset=1&page_id=224 (Conference information)

Conference

Conference	2011 CRCMining Australian Mining Technology Conference
Abbreviated title	Decision Support - embedding process capability in mining systems
Country/Territory	Australia
City	Noosa
Period	12/09/11 → 14/09/11
Other	In an environment where operations are increasingly reliant on technology it isimportant for mine personnel to be given the right tools to make the right decisions. The 2011 CRCMining Australian Mining Technology Conference will explore theplace of technology in all aspects of the mining process, from exploration and mineplanning through to mining methods and materials handling.
Internet address	https://web.archive.org/web/20110723024407/http://www.crcmining.com.au/dynamic_page.php?page_reset=1&page_id=224 (Conference information)

Cite this

Ge, F. L., Moore, W., & Antolovich, M. (2011). Application of learning from demonstration to a mining tunnel inspection robot: The inspection of muck piles and other entities in mines is of keen interest to mining engineers and project managers. This paper presents initial work on developing an automated robot system capable of such inspections. The robot system uses a robot training technique called Learning from Demonstration (LfD), where after human demonstrations a robot can finish a pre-defined task. The paper presents research on decomposing a task into modules that increases robot adaptability in learning an inspection task. The paper shows how modularization is appled to a robot inspecting a muck pile using a Gaussian Mixture Model (GMM) togother with Gaussian Mixture Regression (GMR). Results of simulations and real robot perfromance are presented. In 2011 CRCMining: Australian Mining Technology Conference (pp. 28-35). CRC Mining.

Ge, Feng Lu ; Moore, Wayne ; Antolovich, Michael. / Application of learning from demonstration to a mining tunnel inspection robot : The inspection of muck piles and other entities in mines is of keen interest to mining engineers and project managers. This paper presents initial work on developing an automated robot system capable of such inspections. The robot system uses a robot training technique called Learning from Demonstration (LfD), where after human demonstrations a robot can finish a pre-defined task. The paper presents research on decomposing a task into modules that increases robot adaptability in learning an inspection task. The paper shows how modularization is appled to a robot inspecting a muck pile using a Gaussian Mixture Model (GMM) togother with Gaussian Mixture Regression (GMR). Results of simulations and real robot perfromance are presented. 2011 CRCMining: Australian Mining Technology Conference. Kenmore East, Qld : CRC Mining, 2011. pp. 28-35

@inproceedings{dcad8b121ed34fac9c7408c9e81221d2,

title = "Application of learning from demonstration to a mining tunnel inspection robot: The inspection of muck piles and other entities in mines is of keen interest to mining engineers and project managers. This paper presents initial work on developing an automated robot system capable of such inspections. The robot system uses a robot training technique called Learning from Demonstration (LfD), where after human demonstrations a robot can finish a pre-defined task. The paper presents research on decomposing a task into modules that increases robot adaptability in learning an inspection task. The paper shows how modularization is appled to a robot inspecting a muck pile using a Gaussian Mixture Model (GMM) togother with Gaussian Mixture Regression (GMR). Results of simulations and real robot perfromance are presented.",

author = "Ge, {Feng Lu} and Wayne Moore and Michael Antolovich",

year = "2011",

language = "English",

pages = "28--35",

booktitle = "2011 CRCMining: Australian Mining Technology Conference",

publisher = "CRC Mining",

note = "2011 CRCMining Australian Mining Technology Conference, Decision Support - embedding process capability in mining systems ; Conference date: 12-09-2011 Through 14-09-2011",

url = "https://web.archive.org/web/20110723024407/http://www.crcmining.com.au/dynamic_page.php?page_reset=1&page_id=224",

}

Ge, FL, Moore, W & Antolovich, M 2011, Application of learning from demonstration to a mining tunnel inspection robot: The inspection of muck piles and other entities in mines is of keen interest to mining engineers and project managers. This paper presents initial work on developing an automated robot system capable of such inspections. The robot system uses a robot training technique called Learning from Demonstration (LfD), where after human demonstrations a robot can finish a pre-defined task. The paper presents research on decomposing a task into modules that increases robot adaptability in learning an inspection task. The paper shows how modularization is appled to a robot inspecting a muck pile using a Gaussian Mixture Model (GMM) togother with Gaussian Mixture Regression (GMR). Results of simulations and real robot perfromance are presented. in 2011 CRCMining: Australian Mining Technology Conference. CRC Mining, Kenmore East, Qld, pp. 28-35, 2011 CRCMining Australian Mining Technology Conference, Noosa, Queensland, Australia, 12/09/11.

Application of learning from demonstration to a mining tunnel inspection robot: The inspection of muck piles and other entities in mines is of keen interest to mining engineers and project managers. This paper presents initial work on developing an automated robot system capable of such inspections. The robot system uses a robot training technique called Learning from Demonstration (LfD), where after human demonstrations a robot can finish a pre-defined task. The paper presents research on decomposing a task into modules that increases robot adaptability in learning an inspection task. The paper shows how modularization is appled to a robot inspecting a muck pile using a Gaussian Mixture Model (GMM) togother with Gaussian Mixture Regression (GMR). Results of simulations and real robot perfromance are presented. / Ge, Feng Lu; Moore, Wayne; Antolovich, Michael.
2011 CRCMining: Australian Mining Technology Conference. Kenmore East, Qld: CRC Mining, 2011. p. 28-35.

Research output: Book chapter/Published conference paper › Conference paper › peer-review

TY - GEN

T1 - Application of learning from demonstration to a mining tunnel inspection robot

T2 - 2011 CRCMining Australian Mining Technology Conference

AU - Ge, Feng Lu

AU - Moore, Wayne

AU - Antolovich, Michael

PY - 2011

Y1 - 2011

UR - https://web.archive.org/web/20110723024407/http://www.crcmining.com.au/dynamic_page.php?page_reset=1&page_id=224

M3 - Conference paper

SP - 28

EP - 35

BT - 2011 CRCMining: Australian Mining Technology Conference

PB - CRC Mining

CY - Kenmore East, Qld

Y2 - 12 September 2011 through 14 September 2011

ER -

Ge FL, Moore W, Antolovich M. Application of learning from demonstration to a mining tunnel inspection robot: The inspection of muck piles and other entities in mines is of keen interest to mining engineers and project managers. This paper presents initial work on developing an automated robot system capable of such inspections. The robot system uses a robot training technique called Learning from Demonstration (LfD), where after human demonstrations a robot can finish a pre-defined task. The paper presents research on decomposing a task into modules that increases robot adaptability in learning an inspection task. The paper shows how modularization is appled to a robot inspecting a muck pile using a Gaussian Mixture Model (GMM) togother with Gaussian Mixture Regression (GMR). Results of simulations and real robot perfromance are presented. In 2011 CRCMining: Australian Mining Technology Conference. Kenmore East, Qld: CRC Mining. 2011. p. 28-35

Conference

Other files and links

Cite this