In order for unmanned vehicles such as rovers, aerial vehicles and underwater devices to work in diverse environments and perform previously unknown complex tasks they need to move effectively in diverse terrains and work in cooperation. In order to keep them working properly, it demands these unpiloted conveyance to navigate safely in changing scenarios with unknown static and dynamic obstacles while performing different tasks.

Navigation of mobile robots is a broad topic covering a large spectrum of different technologies and applications. It draws on some very ancient techniques, as well as some of the most advanced space science and engineering. Currently, two relatively modern systems have been extensively studied. One is satellite based Global Positioning Systems (GPS) and the other is image based Vision Positioning Systems, which have the common feature of being under continual development. Between the two, a large scale of navigational requirements can be met. However, these and other techniques have important disadvantages. GPS for example, cannot be used indoors, in underground facilities (e.g. mining operations), tunnels and in environments where GPS information is not readily available. Robots’ motion strategies must rely on sensory information to compute the movements according to the unforeseen circumstances.
Cameras cannot be used in situations when vision is imparred (e.g., fog, smoke, rain, darkness, etc.)

For mobile robotics, lasers, sonars and radios are common media used as navigational beacons. GPS has been widely used in autonomous navigation systems as the navigation sensor replacing the inertial measurement unit, which has been conventionally favored as the primary navigation sensor. Other types of navigational systems employ artificial landmarks.

In order to cope with the many disadvantages of traditional robotic navigational techniques and enable teams of unmanned mobile robots to reach their target in a 3D environment effectively we are investigating a novel navigational 3D method using Intelligent Dynamic-Landmarks (IDL). The methodology is envisioned to improve the maneuvering of diverse types of unmanned vehicles (e.g., ground rovers and aerial drones) while getting rid of traditional navigational errors such as common sensor failures and the potential errors when using dead reckoning.

The methodologies under development have the ability to allow robots to effectively move (navigate) even under the presence of external and internal disturnances such as sensor failure (partial or total).

3D-NUVID:

The focus of the 3D-NUVID project is to develop methodologies to enable teams of unmanned vehicles (e.g., aerial, land rovers, underwater robots, amphibious systems, etc.) to navigate effectively in 3D environments (i.e., land, air, sea or a combination) even under sensor failure.
In order to achieve this we are investigating intelligent cooperating mechanisms, communication protocols, and the use of IDLs. We describe IDL as independent artificial entities (e.g., rovers), with on-board sensors susceptible to failures, that can communicate and move close to each other in order to help other robots determine their exact position. We are investigating how IDLs are used to establish diverse links (e.g., physical and virtual) between all unmanned vehicles comprising the team and how these links can be subsequently used to precisely determine the position of every robot working in a given environment despite sensors and/or robots failure. The navigational approaches under investigation have many advantages over traditional techniques (e.g., can be used in unknown indoors/outdoor environments effectively). We have developed the concept of "Origami Graphs" to track the position and orientation of every robot while providing the tools to compensate for possible errors caused by internal or external disturbances.

Fig. 1: Basic Cooperative Navigation using IDLs.

Fig. 2: Basic Navigation Strategy using IDLs and Origami Graphs.

RESEARCH ISSUES:

Contributions within this project are been made in the following areas: