AUSI is currently undertaking a number of investigations focused on the issues associated with applying autonomous underwater vehicle (AUV) technology to the problems of ocean science. We are part of a much larger research effort focused on developing new ocean sampling systems for the 21st century. This program, "The Autonomous Oceanographic Sampling System (AOSN)" is being funded by the Office of Naval Research and the National Science Foundation and is being undertaken in conjunction with a number of US universities, as well as the Institute for Marine Technology Problems (IMTP), Russian Academy of Sciences. The AOSN concept suggests that networks of mobile and non-mobile autonomous ocean sampling systems can acquire the data necessary to understand large scale ocean phenomena without the use of large and costly research vessels.

Our specific research focus is to investigate the potential of utilizing solar energy to power an AUV such that its endurance will be measured in months as opposed to hours. A solar powered AUV (SAUV) prototype testbed is being developed as one type of AOSN data gathering platform. The solar powered autonomous system will come to the surface each day to recharge its onboard energy system and then undertake ocean data sampling activities during the nighttime hours. While on the surface, it can update its position using GPS and communicate with a remote user to offload acquired data and receive modifications to its onboard instructions. We are specifically focused on the problems that will arise when many of these vehicles are used simultaneously to gather data in a cooperative fashion.

Program Objectives and Description

AUSI and IMTP are working together with the goal of investigating the viability of using an SAUV platform to remotely gather oceanographic data and transmit it to a user on a periodic basis. We have broken down this broad goal into three objectives:

1. To answer questions raised during initial studies and to further develop the SAUV prototype platform
Earlier studies raised questions concerning discrepancies between calculated energy acquired under given wave conditions and actual energy acquired during at sea testing. In addition, the calculated efficiency of the SAUV propulsion system was found to be higher than that measured during at sea testing. This project will focus on clearing up these issues.

Both AUSI and IMTP are currently working on complementary issues of SAUV research using test platforms based on a common vehicle design. These SAUV test platforms will each be modified to incorporate a more modern and capable RTOS environment with the flexibility to utilize TCP/IP communication protocols. It is very important that we have at our disposal a capable and adaptable target system, which is amenable to embedding in current and future vehicle systems. The choice of an RTOS environment is a key step in the long term process of selecting, integrating and testing new sensors and actuators as part of the SAUV program. Such a target environment must be flexible enough to be easily adapted to new (as yet designed) vehicle systems and be powerful enough to take on new computational duties as projects mature and requirements change. Our current goal is to incorporate Octagon hardware board(s) running QNX alongside existing legacy control boards to provide overall control of the legacy system and provide an interface to the vehicle. AUSI may also explore the use of other board configurations in the future, such as PC/104, that support the QNX RTOS.

Finally, we plan to integrate an AUV command language developed from related research, known as Generic Behaviors, into the vehicle control system using the C++ programming language. We will also explore the possibility of using the Metrowerks IDE for cross-platform development when it becomes available.

2. To investigate the issues associated with the remote access and real-time control of an SAUV
AUSI has developed an open and flexible, TCP/IP protocol, client-server simulation environment called the Cooperative AUV Development Concept (CADCON) for development of high-level communication and control for multiple AUVs. CADCON simulation clients are developed in and run on mid-range PC hardware running well-known operating systems, and utilize a simple connection protocol to run over the Internet. We are extending the capability of CADCON to incorporate hardware-in-the-loop. Primary advantages to this approach include (1) the ability to incrementally test many of a real vehicle's subsystems without the accompanying cost and risk associated with an at sea test, and (2) providing an infrastructure for distributing AUV sensor data to interested researchers and for monitoring and control of the vehicle during its mission. As CADCON uses TCP/IP sockets, we would like to preserve this connection mechanism down to the vehicle level, without the requirement for developing a serial interface element. This decision is in line with observed and projected plans of many AUV research groups (as well as research groups across many disciplines) to utilize this well-known protocol in end-to-end system development efforts.

We will define a detailed strategy for the parallel development of SAUV vehicles capable of TCP/IP-based cooperative control. Each of the SAUV systems will support TCP/IP networking, allowing communication between the vehicle and the CADCON environment via the Internet. Issues related to remote monitoring and control, such as bandwidth limitations, packet delays and data frequency requirements, and their impact on CADCON system complexity, will be investigated.

3. To verify the feasibility of using an SAUV to autonomously acquire oceanographic data and relay that data to a remote user on a regular basis by conducting a 30 day autonomous data gathering experiment
We will establish operational scenarios to account for the environmental parameters impacting SAUV system performance and determine how closely actual performance matches the calculated performance. We will work on defining appropriate sensors for the mission and integrating the sensor suite onto the SAUV. The mission will be conducted over a 30 day period, with 12-24 hour updates from the vehicle when it surfaces to recharge its batteries. Following the mission, we will verify the quality of the SAUV acquired data by comparison with similar data acquired by conventional sampling methods.