Research

RESEARCH FOCI:

The research foci of the More Than One Robotics laboratory are in intersection of three research themes: Mobile Sensor Networks, Modular Robotics and Human-Robot Cooperation. 


We exploit communication and interaction of human and robotic agents to develop systems theory and algorithms for the complex intelligent systems that can be evaluated through mathematical modeling, simulation, real-world experiments. 





CURRENT PROJECTS: 

(to be updated soon, visit our youtube for several video demonstrations of our research projects


DRAMA: Deformable Group of Affordable Robots for Multi-modal Cleaning (BRC fund)
PI: Trung Dung Ngo
Co-PI: Co-PI: David James Young, Lim Lee Hoon, Voo Nynk Yoong, Hartini Binti Hj Mohd Yasin, Shaoping Bai, Henrik Schiøler, John Leth, Vang Le Quy

While the labour cost is rapidly increasing all over the world, service robots become the enabling technology to reduce the labour cost by releasing human workers out of simple jobs. Low-cost cleaning robots are expected to replace cleaners in various domestic and professional environments. However, cleaning robots have widely used yet because the major barrier is the low level of cleaning quality and coverage efficiency. New robotic technology is, therefore, urgently demanded to improve the performance of cleaning servicesIn this three-year project, we aim to develop a deformable group of affordable cleaning robots for multi-modal cleaning tasks. The primary objective is to enable a group of cleaning robots to work collaboratively with multiple modes adaptable to workspaces. The affordable robots are upgraded to highly controllable and flexible levels by deformation and coverage planning algorithms. The project will contribute to the development of next generation cleaning robots working in large-scale workspaces: (1) reduce time consuming and cost of system installation; (2) improve cleaning performance in terms of cleaning quality and coverage; (3) increase flexibility of the cleaning robots for different environments

SENSAFE: Human Safety in Human-Robot Cooperation ( UBD/PNC2/2/RG/1(259))
Researchers: Tung Xuan Truong, Trung Dung Ngo (PI)
-------------------------------------------------------------------
A challenge of applying mobile robots to our daily life is safety: how to ensure that robots will reliably assist human, not harm them. In SENSAFE we suppose that human and service robots share the workspace, and implicitly or explicitly cooperate to finish assigned tasks. Risks of human in the workspace essentially come from the attack of robots caused by its functioning failures or misunderstanding between human and the robots. Based on the general close-loop-model of robotic 
systems, “sense-think-act”, “sense” is the first component that plays a role as perceptual, representational, and reasoning input to the sequence “think-act”, thus “sense” is the essential reason causing the malfunction of the robots due to uncertain conditions of the working environment and unpredictable actions of human co-workers.
         We propose a research method, called as experiment-based knowledge (empirical knowledge) on how to determine modes of operation and control of robots according to safeguarding and clearance between the robots and human in the human-life workspace based on the robot’s “sense”. The objective of SENSAFE is to develop experiments for assessing safety levels of human co-working in the shared workspace. The outcome of this project is to contribute to: 1) society in terms of deeper understanding of human risks in the manner of human protection and injury minimization when sharing workspace with service robots; 2) industry in terms of technological standards for development of service robots working in human shared workspace; 3) science in terms of techniques and algorithms for safe human-robot interaction/cooperation


Two-stage Optimisation of Communication Throughput in Large-scale Mobile Sensor Networks
Researchers: Adnan Fida, Pg. Jaidi Tuah, Trung Dung Ngo (PI)
-------------------------------------------------------------------
We concern the development of algorithms for optimising the network topology and the information link capacity of large-scale mobile sensor networks. The bottlenek of the network must be found for the cooperative control of node mobility leading to improvement of the network capacity and communication throughputs. Both centralized and decentralised solutions are developed for comparative analysis. A simulator is built for implementation and validation of the proposed algorithms. The real-work experiments will be conducted on the mobile robot platforms available in the lab for real data analyses and demonstrations. 

1. T.D.Ngo, LindMind: Link Optimisation of Swarming Mobile Sensor Networks, Sensor Journal, 2011 (ISSN 1424-8220).
2. T.D.Ngo, Distributed Optimization of Communication Capacity of Swarming Mobile Sensor Networks. the IEEE Int. Conference in Ubiquitous Robot and Ambient Intelligence, Deajeon, Korea, Nov.2012. 

3. A.Fida, T.D.Ngo. A Survey on Routing Protocols of Mobile Sensor Networks, in preparation. 
4. T.D.Ngo, Two-Stage Optimisation of Communication Throughput for Mobile Sensor Networks, the 13th Int. Conference in Intelligent Autonomous Systems, Padova, July 14-19, 2014 (under review)

   

A Decentralised Control of Robot Formation and Transformation
Researchers: Muhammad Iqbal, Pg. Jaidi Tuah, John Lelf, Trung Dung Ngo (PI)
-------------------------------------------------------------------

We develop mathematical modelling and control for achieving different formations and the transformation of multi-agent robotic systems. The developed mathematical modelling and control algorithms provide high flexibility with a decent amount of achievable formations. Moreover, this method allows the robot formation to transform their formation shapes with and without sensing, perception, and communication capabilities according to the environment conditions. 


1. M. Iqbal, T.D.Ngo, On the Formation Control of Multiple Agents under Nearly Cyclic Pursuit, the 13th Int. Conference on Autonomous Intelligent Systems, Padova, Italia (under review). 


Inter-communcation and Routing of Reconfigurable Robots
Researchers: Tung Van Le, Trung Dung Ngo (PI)
-------------------------------------------------------------------
Routing and intercommunication is the crucial factor for large-scale modular robotic systems. To allow the robotic system reconfigurable for various applications, the robotic building blocks must guarantee uninterrupted intercommunication through routes of intelligent building blocks. Smart routing protocols for intercommunication for reconfigurable systems will be investigated and validated through simulation and real-world building blocks.


Self-deployment Strategies of Multi-agent Robotic Systems in Unknown Environments
Researchers: Hung Duy Pham, Vinh Quang Tran, Trung Dung Ngo (PI)
-------------------------------------------------------------------

This project addresses a decentralized control of robot swarm for multi-target tracking with network preservation in unknown environments. The robot swarm moves towards targets while preserving connectivities for network communication. The developed controller relied on local information of neighboring robots is synthesized by the rules of behavioral control and connectivity maintenance. The swarm dispersion algorithm based decentralized control consists of three functionalities: (1) maintaining connectivities between the robots for swarm movement towards the targets; (2) breaking down swarm topology to accelerate the target reaching; (3) adjusting the robots’ velocity to preserve critical connectivities for a communication network of the robots. The developed algorithm is demonstrated and verified in simulation and real experiments. 

1. H.D. Pham, T.D.Ngo (2013), Accelerating to Multi-Target Tracking by A Robot Swarm with Network Preservation, IEEE Int. Conf. Soft-computing and Pattern Recognition, Hanoi, Vietnam, Dec 15-18, 2013.
2. H.D.Pham, T.D.Ngo, Scalability of a Large-Scale Swarm of Mobile Robots with End-to-End Network Preservation for Multi-target Tracking, the 13th Int. Conference in Intelligent Autonomous Systems, Padova, July 14-19, 2014 (under review).


Ants on a Stick: Robotics Education for Any Level
Researchers: Tung Xuan Truong, Tung Le Truong, Adnan Fida, Muhammad Iqbal, Trung Dung Ngo (PI)

Inspired from the primary school mathematical problem about ants on a stick, we are developing a system of mobile robots that mimic ants operations. This project aims at educating students the meaningful roles of learning mathematical theories and physics laws for their future higher education. On the other hand, this project is to promote the university education through series of demonstrations on the real systems. 


Problem-based Learning in Robotics Education
Researcher: Trung Dung Ngo
-------------------------------------------------------------------
We consider technical aspects using Problem-Based Learning (PBL) in Robotic Education. As robotics is an interdisciplinary subject that require students having theoretical and practical skills, using PBL method provides students highly potential possibilities to develop their learning outcome through various learning activities, e.g., teamwork skills, hands-on practice skills, problem solving skills, and project planning and organization skills. 
We carry out our research by conducting different robotic projects for the students' semester theses or final year thesis in the PBL atmosphere and observing their learning progress and outcome. 

1. T.D.Ngo, Using Multiple Mobile Robots as an Educational Tool for Problem Based Learning, In preparation
2. T.D.Ngo, Enhancing Early Stage Researcher in Robotics by Project Organised Problem-Based Learning, In preparation




COMPLETED PROJECTS:


Bio-inspired Heterogeneous Swarm of Mobile Robots for Fast Deployment and Exploration
-------------------------------------------------------------------
We develop a swarm of mobile robots can be applied in many applications including exploration and mapping, surveillance and reconnaissance, patrol and monitoring, victim identification in hazardous environments. More advantageous than a single robot, a robot swarm should be sent into human impassible areas for cooperative exploration and data gathering. Such emergency response services require a strategy allowing fast deployment and exploration of mobile robots.
This work presents a strategic method and technological solution to emergency response services by inspiration from the Kangaroo family. We describe a semi-autonomous system consisting of a Kangaroo “mother” robot carrying a number of “child” robots sitting on her “pouch”. Depending on situation, the mother allows the children to jump out exploration, data gathering, or call them back for data collection and displacement. 


1. T.D.Ngo, A Hierarchical Mobile Sensor Network for Fast Deployment and Exploration in Human Impassible Environments, in proceedings of the IEEE International Conference on Ubiquitous Robots and Ambient Intelligence, Deajeon, Korea, Nov 26-28, 2012 (the Best Video Award)
2. T.D.Ngo (2014), Kangaroo-Inspired Robotic System: A System of Systems to Assist Fire Fighters, The 19th RSJ Int. Symp. Artificial Life and Robotics, Beppu, Japan, Jan 22-24, 2014.
3. T.D.Ngo (2012), A Hierarchical Robotic System for Search and Rescue, Video Presentation at AAAI-12, Toronto, Canada, July 22-26, 2012
4. T.D.Ngo (2011), A Swarm of Kangaroo robots for Emergency Responses, presented in European Network for the Advancement of Artificial Cognitive Systems, Interaction and Robotics, Thessaloniki, Greece, April 11-12, 2011  
5. T.D.Ngo, Kangaroo Robots: Versatility of A System of Systems in Robotics, Int.t Journal of Bionic Engineering, 2013 (under review)
 

Dance with a Swarm of Robots
-------------------------------------------------------------------
Dancing is one of our daily life activities in which a couple or a group can perform the art of body acting. Dancing can be seen as cooperative action in which one acts under combination of “passive” and “active” constraints. That mean, the constraint is “passive” if all are acting with respect to a central signal, e.g., musical rhythm while the constraint is named as “active” if one act to mimic or response to the behaviour of their partner/neighbours. However, dancing with robots is very rarely seen in our life.

In this project, we are going to make an intelligent interface for human robot interaction with the embodiment of  dancing. The hypothesis of this project is about a mixed swarm of mobile robots and human dancing together. In the case, human is the factor creating “digital” musical rhythm for robots while the robots automatically maintain their attitudes by observing their neighbors. The research challenges are about distributed computing of robot swarm and wireless network of human-robot interaction. The target of this project is the protocol of a network of distributed computing units acting under human control synchronously and the algorithm maintaining the robot attitudes in the swarm. 

1.  T.D.Ngo, Human-robot Swarm Interaction: A Case study of Dancing with Robot Swarm, In preparation

A Simplified Control of Robot Convoy
-------------------------------------------------------------------

We study a method on how to develop a simplified control of robot convoy. We approach the problem with a control based on multi-level “bang-bang” by integration of Sensory Input Map (SIM) and Control Output Map (COM). The developed control is similar to behaviour-based control because it is a set of rules that are derived by mapping sensory data into motor values. We demonstrate that the robots are able to follow the leader while keeping a safe constant distance between them. The developed control is validated in simulation and tested with up-to five mini-robots. 



1. H.D.Pham, T.D.Ngo (2012), A Simplified Control of Robot Convoy, in proceedings of International Conference on Control, Automation, and Information Science, Saigon, Vietnam, Nov 26-29, 2012

Probabilistic Communication based Potential Forces for Robot Formation
-------------------------------------------------------------------

We introduce a new method of artificial potential forces based on probablistic communication, called (Probablistic Communication based Potential Forces - PCPF). The potential forces provide a locally distributed control for a large volume of self-regulated formation of mobile robots. While models of sensing and communication have mostly existed with simple assumptions that are far away from the physical properties of sensors and communication mechanisms, the method is realistic because both attractive and repulsive forces are only based on probability of communication which are empirically measured and approximately estimated between robots. The method is demonstrated through various non-trivial examples of robot formation and formation transformation. Analysis are provided to facilite understanding of the elements of the probablistic method.

1. R. Jespersen, S.B.Mikkelsen.B.S, T.D.Ngo, A unified framework for swarm algorithm development and evolution, in proceedings of IARP workshop on Robots for Risky Interventions and Environmental Surveillance – Maintenance, January 2010, 20-21, Sheffield, UK 
2. S.B.Mikkelsen, R. Jespersen, T.D.Ngo (2010), Probabilistic Communication Based Potential Force for Robot Swarm: A Practical Approach, in proceedings of the 10th International Symposium on Distributed Autonomous Robotics Systems, Lausannse, Switzerland, November 1-3, 2010

Communication-based Decision Making in Robot Swarms
-------------------------------------------------------------------
We present collective decision making as the result of  emergent self-organization, in which individuals allow to contribute to group decision. Benefits of collective decision are able to make group consensus of rapidly carrying out united mission without power struggles of leadership election. Social insects are broadly known with emergent self-organization for making collective decision of carrying out missions beyond the capabilities of individuals: ants collaborate to carry sugar grains exceeding the body weight of individuals, or termites build huge mounds. Swarm robotics is a biologically inspired field which inherits significantly from the observation of biological populations such as social insects like ant colonies, termites, bees, and wasps, inspiring the vision of how a large number of simple individuals can collectively collaborate with other members to release intelligent systems. The hypothesis of this project is about a swarm of mobile robots that can issue some collective decision based on limited capability of sensing and communication. The research challenges are about directional wireless signal based collection decisions for several applications such as swarm movement, collectively searching and detecting victim. The target of this project is bio-inspired control for single robots and collective algorithms for the swarm.

1. H. Schiøler, T.D.Ngo, Epidemic Synchronization in Robotic Swarms, in proceedings of IFAC International Workshop in Networked Robotics (NetRob09), Colorado, USA, October, 06-08, 2009 
2. T.D.Ngo (2010), Quantitative Analysis of Distributed Control Paradigms for Robot Swarms, in the proceedings of the IEEE international conference on Robotics and Biomimetics, Taianjin, China, December 14-18, 2010



Towards Energy Autonomy in Multi-agent Robotic Systems
-------------------------------------------------------------------
We study aspects of self-sufficient energy (energy autonomy) for truly autonomous robots and towards sociable robots. Over sixty years of history of robotics through three developmental ages containing single robot, multi-robot systems, and social (sociable) robots, the main objective of roboticists mostly focuses on how to make a robotic system function autonomously and further, socially. However, such approaches mostly emphasize behavioural autonomy, rather than energy autonomy which  is the key factor for not only any living machine, but for life on the earth. Consequently, self-sufficient energy is one of the challenges for not only single robot or multi-robot systems, but also social and sociable robots. This thesis is to deal with energy autonomy for multi-robot systems through energy sharing (trophallaxis) in which each robot is equipped with two capabilities: self-refueling energy and self-sharing energy. In addition to behavioural autonomy, the robots obtaining the capabilities are able to become truly autonomous and further sociable robots. 

1. T.D.Ngo, H. Raposo, H. Schiøler.(2008). Multi-agent Robotics: Towards Energy Autonomy, International Journal of Artificial Life and Robotics (AROB), Springer Publisher, Vol. 12, No. 1-2, pp.47-52 
2. T.D.Ngo, H.Schiøler.(2008).Randomized Robot Trophallaxis, Chapter in the Book of Recent Advances in Multi-robot SystemsISBN 978-3-902613-24-0, pp.197-232

3. H. Schiøler, T.D.Ngo (2009), Epidemic Synchronization in Robotic Swarms, in proceedings of IFAC International Workshop in Networked Robotics (NetRob09), Colorado, USA, October, 06-08, 2009
4. T.D.Ngo, H. Schiøler, Rendezvous Trajectory Generation for Energy Trophallaxis, in proceedings of IEEE conference in Control, Automation, Robotics and Vision (ICARCV2008), Hanoi, Vietnam, December 17-20, 2008.  

5. H. Schiøler, T.D.Ngo (2008), Trophallaxis in Robotic Swarms – Beyond Energy Autonomy, in proceedings of IEEE International Conference in Control, Automation, Robotics and Vision (ICARCV2008), Hanoi, Vietnam, December 17-20, 2008 (invited paper).
6. T.D.Ngo, H. Schiøler (2007), Randomized Robot Trophallaxis: From Design to Implementation, in proceedings of IEEE/SMC International Conference in Systems, Man, and Cybernetics (SMC07), Montreal, Canada, October 07-10, 2007
7. T.D.Ngo, H. Raposo, H. Schiøler (2007). Being Sociable: Multirobot with Self-sustained Energy, in the proceedings of the 15th Mediterranean Conference on Control and Automation (MED’07), Athen, Greece, June 27 – 29, 2007
8. T.D.Ngo, H. Schiøler. (2007). Potentially Distributable Energy: Towards Energy Autonomy in Large Population of Mobile Robots, in proceedings of The 7th IEEE International Conference on Computational Intelligence in Robotics and Automation (CIRA'07), Jacksonville, Florida, USA, June 20 -26, 2007
9.  T.D.Ngo, H. Raposo, H. Schiøler. (2007), Multi-agent Robotics: Towards Energy Autonomy, in proceedings of 12th International Conference in Artificial Life and Robotics, Oita, Japan, January 25-27, 2007. (the Best Student paper).

10. T.D.Ngo, H.Schiøler. (2006) Sociable Robots through Self-Maintained Energy, International Journal of Advanced Robotic Systems (ARS), Vol.4, pp.313-322
11. T.D.Ngo, H. Schiøler. (2006). An Approach to Sociable Robots through Self-distributed Energy, in proceedings of IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), Beijing, China, Oct 9-15 2006 
12. T.D.Ngo, H. Schiøler. (2006). Probabilistic Distributed Energy for Long-lived Group of Mobile Robots, in proceedings of IEEE International Conference on Cybernetics & Intelligent Systems, and Robotics, Automation, & Mechatronics (CIS-RAM), Bangkok, Thailand, June 7-9 2006



Tangible Programming by Modular Robotics
-------------------------------------------------------------------
We investigate different aspects of modern artificial intelligence using modular robotics. The intelligent building blocks are robotic modules that are possibly used to develop various applications through the concept "Tangible Programming". We proposed information theories applied to build a number of distributed modular systems using intelligent building blocks. To illustrate these possibilities, we present a range of differently experimental scenarios in which intelligent building blocks have been used to set-up pervasive robotic applications, location-aware computing system, and even modular wireless sensor network.

1. T.D.Ngo, H.H.Lund. (2004). Modular Artefacts, Position paper for ECOOP 2004 workshop: Component-oriented Approaches to Context-aware Computing, Oslo, Norway, June 14th, 2004
2. T.D.Ngo, H.H.Lund. (2004) Ambient Intelligence in Distributed Modular Systems, in proceedings of  the 2004 International Conference on Electronics, Information and Communications ICEIC 2004, Hanoi, Vietnam, August 16th-18th, 2004. Vol1. Page.410-416
3. T.D.Ngo. (2005). Modern Ambient Intelligence Embodied in Distributed Modular Robotic Systems, in proceedings of The 3nd International Conference on Computational Intelligence, Robotics, Autonomous Systems - CIRAS, Singapore, December 13-16, 2005.·