Danilo Tardioli

Enforcing Network Connectivity in Robot Team Missions

The growing interest in robot teams for surveillance or rescue missions entails new technological challenges. Robots have to move to complete their tasks while maintaining communication among themselves and with their human operators, in many cases without the aid of a communication infrastructure. Guaranteeing connectivity enables robots to explicitly exchange information needed in collaborative task execution, and allows operators to monitor or manually control any robot at all times. Network paths should be multi-hop, so as not to unnecessarily restrict the team’s range. In this work we contribute a complete system which integrates three research aspects, usually studied separately, to achieve these characteristics: a multi-robot cooperative motion control technique based on a virtual spring—damper model which prevents communication network splits, a task allocation algorithm that takes advantage of network link information in order to ensure autonomous mission completion, and a network layer which works over wireless 802.11 devices, capable of sustaining hard real-time traffic and changing topologies. Link quality among peers is the key metric used to cooperatively move the robots and maintain uninterrupted connectivity, and the basis for novel ideas presented in each subsystem. Simulations and experimental results with real robots are presented and discussed.

Real Time Communications over 802.11: RT-WMP

Ad-hoc networks usually support best-effort traffic and occasionally some kind of quality of service (QoS). However, there are some applications, which generally involve cooperative control, with hard real-time traffic requirements where strict deadlines must be met. To meet deadlines, the communication network has to support the timely delivery of inter-task messages. This is the case, for example, of applications involving cooperative robot teams, such as those used for rescue tasks in hostile environments, emergencies or disaster recovery, where a wired backbone is in-feasible or economically unviable. In this paper, we present RT-WMP, a novel protocol that allows wireless real-time traffic in relatively small mobile ad-hoc networks using the low-cost commercial 802.11 technology. The protocol is based on a token-passing approach and message exchange is priority based. Moreover, support for frequent topology changes is provided through the sharing of a matrix that describes link quality amongst the members of the network.

Real-time wireless multi-hop protocol in underground voice communication

The underground communication in tunnels and mines is very challenging due to the hostile nature of the environments and to the propagation issues that electromagnetic waves suffer there. Communication is often unidirectional (e.g. in mines) or very costly (e.g. leaky feeder in road tunnels) and hard to install and maintain. This work proposes the use of multi-hop ad-hoc networks to provide multimedia communication between mobile nodes in such a hostile environments, relying on a complete hardware/software, cheap and easy-to-setup platform that can be used both as temporary or fixed infrastructure or as communication backbone in emergency scenarios like mine accidents or a tunnel collapse. The communication is based on the Real-Time Multi-hop Protocol (RT-WMP) and its QoS extension executed over several nodes equipped with specific hardware. This protocol manages delay sensitive messages and the node mobility across the network while the QoS extension is responsible for allowing the end-to-end voice communication. The specific topology and situation have driven to a specialization of RT-WMP to better perform in this type of environments, taking advantage of the a priori (partial) knowledge about the topology. This proposal was tested in a real application in the Somport tunnel, the about 8km-long railroad linking Canfranc, Spain with Pau, France.

A wireless multi-hop protocol for real-time applications

Mobile Ad-hoc NETworks (MANETs) have been gaining increasing popularity in recent years thanks to their ease of deployment and the low cost of their components. The routing protocols in ad hoc networks face the challenge of establishing and maintaining multi-hop routes while complying with mobility, bandwidth limitation and power constraints. The already demanding problem of offering wireless communication in a MANET becomes more complicated in the case of real-time systems where the loss or late arrival of a single item of data can cause serious problems. In this article, we propose a real-time wireless protocol for MANET capable of timely delivery of data. Taking advantage of a cross-layer design, it includes a novel medium access control mechanism and routing algorithm based on the link-quality among the nodes belonging to the network. The protocol manages message priority and is capable of multi-hop communications. It has been conceived mainly to provide real-time wireless communication for small robot teams, making possible the sharing of information such as kinematics or laser data. The validity of the protocol is proven by an in-depth theoretical analysis of its real-time characteristics and performance and through a set of real-world experiments.

Robot Teams for Intervention in Confined and Structured Environments

Safety, security, and rescue robotics can be extremely useful in emergency scenarios such as mining accidents or tunnel collapses where robot teams can be used to carry out cooperative exploration, intervention, or logistic missions. Deploying a multirobot team in such confined environments poses multiple challenges that involve task planning, motion planning, localization and mapping, safe navigation, coordination, and communications among all the robots. To complete their mission, robots have to be able to move in the environment with full autonomy while at the same time maintaining communication among themselves and with their human operators to accomplish team collaboration. Guaranteeing connectivity enables robots to explicitly exchange information needed in the execution of collaborative tasks and allows operators to monitor and teleoperate the robots and receive information about the environment. In this work, we present a system that integrates several research aspects to achieve a real exploration exercise in a tunnel using a robot team. These aspects are as follows: deployment planning, semantic feature recognition, multirobot navigation, localization, map building, and real-time communications. Two experimental scenarios have been used for the assessment of the system. The first is the Spanish Santa Marta mine, a large mazelike environment selected for its complexity for all the tasks involved. The second is the Spanish-French Somport tunnel, an old railway between Spain and France through the Central Pyrenees, used to carry out the real-world experiments. The latter is a simpler scenario, but it serves to highlight the real communication issues.

A wireless communication protocol for distributed robotics applications

Wireless communication is necessary in any multi-robot application which is carried out in the real world. Sometimes a single access point that connects all the units is enough but often support for multi-hop communication is needed to avoid the mobility limitations that centralized routing entails. However, common communication protocols do not have native support for multi-hop communication and must rely on upper layer routing protocols. Nevertheless, these protocols have not been designed to manage high mobility nor flows priority and often offer insufficient performance in scenarios with this kind of requisites. In this paper we present a wireless protocol specifically designed for communication in distributed robotics systems and capable of managing high mobility of nodes while maintaining good performance in terms of fairness and end-to-end delivery delay.

Distributed implementation of discrete event control systems based on Petri Nets

In this paper, we propose a platform for the implementation of Distributed discrete event control systems. We propose an architecture for the Distributed implementation of Petri Nets (PN) in control applications. The architecture covers the local execution of PN and the communication amongst controllers. Both synchronous and asynchronous communication paradigm is supported. A framework of classes for supporting the real time communication and real time execution or remote calls has been developed. The Real Time Wireless Multi-hop Protocol has been adopted. With this protocol and the framework of classes developed, control deadlines of the system can be guaranteed. In order to solve access to common resources, a new class of communications places, the Auction Communication Places, is proposed. To support the real time and concurrent characteristics, we used the Java real-time specification. In order to implement PN, we have extended the concurrent coordinators technique, developed in previous works, to distributed systems introducing the distributed coordinators. To demonstrate the practical utility and feasibility of the architecture, we have applied it to the control of a flexible manufacturing cell.

Spatial diversity based coverage map building in complex tunnel environments

Knowing certain characteristics about signal behavior in specific environments is crucial in order to perform effective utilization of a wireless network, and one manner to achieve this is through the employment of coverage maps. We designed a tool for coverage map building, based on various modules, and implemented them over a real all-terrain vehicle and a Pioneer P3AT robot. We test both tools in the Somport railway tunnel, for which a specific propagation model is unavailable due to characteristics such as lateral vaults and galleries, horseshoe shape, and a change in slope, among others. Results were compared to a Straight Tunnel Ray Tracing model. We found similarities and differences regarding fadings. Also, we were able to dismiss significant influence of vaults and lateral galleries over propagation. We found that due to transversal variations, good quality signal can be achieved in almost the entire studied area by applying a large scale diversity for antennas, and coverage maps were constructed on this basis. Results derived from this work can be used from optimal deployment of a wireless network, up to the development of multi-robot navigation strategies under communication constrains in this type of scenarios.

Adding multicast capabilities to wireless multi-hop token-passing protocols: Extending the RT-WMP

Support for real-time traffic and multi-hop is a basic requirement in certain applications as, for example, cooperative robot team missions. Normally, point-to-point communication is sufficient to allow such a collaboration. However, there exist situations in which multicast and/or broadcast capabilities, even with real-time requirements, allow a better usage of the available bandwidth leaving more time for unicast communications. This is the case, for example, in which a server generates real-time traffic that has to be delivered to a subset (that can be the whole set) of the nodes of the network. In this paper we present a solution to incorporate the capability of sending multicast messages in token-passing real-time wireless protocols. The solution has been implemented and tested as an extension of RT-WMP (a real-time wireless multi-hop protocol with priority management support). In addition, using these new characteristics, another type of unicast/multicast protocol is proposed, analyzed, evaluated and compared with the plain RT-WMP.

A resource allocation strategy in a robotic ad-hoc network

Ad-hoc networks, being infrastructure-less, have much scarcer amounts of bandwidth. Most multi-robotic applications adopt decentralised approaches in order to avoid single points-of-failure. A decentralised approach would require robots to continuously update each other to coordinate their actions. For successful coordination, hard real-time requirements must be met. Thus, a resource allocation strategy is required to fulfil these requirements. This paper presents a novel resource allocation strategy that uses auction-based methods to allocate the network resource among nodes in a robotic ad-hoc network. The strategys aim is two-fold: 1) to allocate the network fairly among all nodes when unmanned; 2) to allocate more of the network resource to a node that the user may favour over the other nodes. It has been tested in two different indoor environments to see how environmental factors affect the performance of the system. The experimental results show that the proposed resource allocation strategy improves the use of bandwidth and enhances transmission.