Christophe Guettier

Automatic vehicle navigation with bandwidth constraints

This paper introduces a new approach to manage network access in tactical environment for specific vehicles. These vehicles must ensure continuity of communications on the move with the command chain and with other units, like dismounted soldiers. Due to tactical mobility, path loss and versatile threats, vehicle navigation must be managed online according to communication connectivity, mission and operational constraints. The paper presents an on-line planning technique that optimizes navigation plans according to available communications. The navigation problem is modelled using constraint formulations to express both connectivity and operational requirements. Navigation plans are solved and optimized according to a combination of a metaheuristic and constraint solving techniques. Various forms of optimization strategies are compared and evaluated on realistic scenarios.

A constraint-based approach for planning unmanned aerial vehicle activities

Unmanned Aerial Vehicles (UAV) represent a major advantage in defense, disaster relief and first responder applications. UAV may provide valuable information on the environment if their Command and Control (C2) is shared by different operators. In a C2 networking system, any operator may request and use the UAV to perform a remote sensing operation. These requests have to be scheduled in time and a consistent navigation plan must be defined for the UAV. Moreover, maximizing UAV utilization is a key challenge for user acceptance and operational efficiency. The global planning problem is constrained by the environment, targets to observe, user availability, mission duration and on-board resources. This problem follows previous research works on automatic mission Planning & Scheduling for defense applications. The paper presents a full constraint-based approach to simultaneously satisfy observation requests, and resolve navigation plans.

Interoperable joint planning and execution Web service with TITAN

Defense systems are involved nowadays in missions that require a high level of interoperability. The paradigm of software oriented architecture (SOA) is generally presented as an efficient approach to tackle the lack of technical interoperability between joint tactical land systems. This paper present a practical experiment, between representative US and France chain of command, that focuses on a collaborative mission planning Web Service (WS). The operational expectation is to improve mutual understanding between allied forces to better perform assistance requests between ground supports (logistic, medevac, …) and to improve coordination between combat units. The mission planning service has been implemented through Web technologies like SOAP, UDDI and XML Schema and relies on the CERDEC TITAN Workflow Orchestration Service. It also uses proactive and reactive planning computation to analyze Friendly Course of Action (FCOA) at preparation time as well as during mission execution. Hosted by the C4ISR OTM team, contributions and main results of the experiments are detailed. Results highlight the benefit of WS design in spite of highly complex functionalities, the interest for automated functions encapsulation and also assess shared awareness involvement in the joint tactical planning process.

Automated planning and scheduling service for Network based Operations

Future Network based Operations (NbO) will necessitate new approach to deal with assymetric context, urban area and tactical mobility. These operations will have to cope with multiple collaborative actions in a complex and changing environment. In NbO, collaborative actions stresses coordination and synchronisation requirements, both in terms of complexity and tempo. Consequently, NbO deeply impacts plan elaboration and task scheduling processes performed by the different units. This paper presents a constraint-based method for automating mission planning and scheduling services in the context of Network based Operations. The approach is experimented on realistic benchmarks and during the Phoenixpsila07 experiment. Results show that mission planning and scheduling tools can be used to develop, experiment and evaluate the NbO concept.

Automatic Optimisation of Reliable Collaborative Services in OLSR Mobile Ad Hoc Networks

Future Network based Operations (NbO) will strongly rely on Mobile Ad hoc Networks (MANET), due to urban area, tactical mobility and assymetric operation management. These networks will support multiple collaborative services, such as sensor to shooters, reachback, maintenance of Common Operational Picture (COP). Both networks and services will have to be managed with no or limited operator intervention, still providing reliable behavior in spite of aggressive environments. At routing level, we present how to preserve 2-connectivity by adapting the Optimised Link State Routing Protocol (OLSR). We also introduce the concept of active subnet management to retrieve maximal operational gain from collaborative services. Following a constraint solving method, the paper shows how to maximise the subnet of actors, while satisfying 2-connectivity, reactivity and communication quality constraints. We demonstrate the approach on simulating real world NbO.

A Consensual and Non-ambiguous Set of Gestures to Interact with UAV in Infantrymen

In the context of using an Unmanned Aerial Vehicle (UAV) in hostile environments, gestures allow to free the operator of bulky control interfaces. Since a navigation plan is defined before the mission, only a few commands have to be activated during the mission. This allows a gestural symbolic interaction that maps commands to a set of gestures. Nevertheless, as gestures are not universal, this asks the question of choosing the proper gestures that are easy to learn memorize and perform. We propose a four step methodology for eliciting a gestural vocabulary, and apply it to this use case. The methodology consists of 4 steps: (1) collecting gestures through user creativity sessions, (2) extracting candidate gestures to build a catalogue, (3) electing the gesture vocabulary and (3) evaluating the non-ambiguity of it. We then discuss the relevance of the GV.

Development and experimentation of collaborative Red Force Tracking in service oriented architecture for tactical networking systems

This paper presents the development and the experimentation of new collaborative services in order to manage global tactical information. The aim is to improve perception and awareness of the surrounding events at the tactical level in order to improve mission execution, bringing higher safety and allowing quicker actions for protection. These services are implemented as distributed processes over tactical information systems. The paper describes service urbanisation that encapsulates automatic tracking, course of action analysis, user interactions and related interfaces. A particular insight is provided on data fusion functionality. It generates hypothesis on spatial and temporal observation associations, on different time scales to provide a relevant Enemy Course of Action. The service characterises targets and estimates tracks but also performs consistent checking over the set of tracks and observations. Results have been provided in the frame of the Phoenix 2010 experiment.

Demonstrating new tactical capabilities with heterogeneous network-enabled systems: The Phoenix 2007 experimentation

In network based operations (NbO), critical decisions can be optimised thanks to a better information sharing between commanders. The Phoenix 2007 experimentation applies the NbO concept to the platoon level. The experimentation introduces collaborative services for video sharing and decision making in order to cope with urban area, tactical mobility and uncertain environment. They open new spectrum of capabilities (short loops between actors, experts reachback, remote navigation for drones or robots...), based on communication and information advantage. For the experimental purpose, these services rely on a system of tactical networks which involve remote sensors, soldiers, vehicles, robots, drones, and expert reachbacks. A Information Support Team (IST) is also introduced to filter and route the traffic flows towards end users. This team is also in charge of deploying and controlling sensors, robots and drones. The experimentation uses basic infantry missions to assess these capabilities, by measuring coordination between actors, traffic loads on the network as well as ergonomic factors. Results reveal that even if robotic sensors and drones do not slow down the pace of operations, good coordinations combined with a versatile IST commanding are required. At system level, network mobility and robustness are required to operate efficiently.

Partial Automatization of Legacy Systems Integration Using Web Application Creator

Interoperability is as old as networked applications. New contexts and technics offer new challenges and new opportunities to cope with them. This paper presents a way to use modern technics and C2-SENSE concepts of Profile to partially automatize the integration of existing systems into a standardized framework used for interconnection of emergency applications. This approach has been proven as efficient in C2-SENSE pilot application.

Hybrid solving technique for vehicle planning with communication constraints

The complexity of modern engagements and the numerous communication links made available increase the difficulties of mission planning and execution. In particular, finding unit waypoints while meeting frequencies allocation is a hard operational problem. Continuity of communications among the command chain must be guaranteed on the move in spite of tactical mobility, networks availability, path loss, jamming, mission updates and versatile threats. In previous works, we proposed a problem formulation relying on vehicle plan optimization, that satisfies network connectivity and operational constraints. This paper adress hybrid search techniques to solve Bandwidth-constrained Vehicle Planning (BVP) on-line problem instances. It is modeled using constraint-based formulations to express both network connectivity and operational requirements (missions, waypoints, terrain). Vehicle planning is then solved and optimized according to a combination of a metaheuristic, namely an Ant Colony Optimisation (ACO), and Constraint Programming (CP) techniques. Various forms of optimization strategies are compared and evaluated on realistic scenarios, outperforming previous results.