AUTOMATION AND CONTROL

Introduction

Automation and control are important technologies for underwater vehicles. Automation can pertain to a range of capabilities, from simple systems that assist the operators in performing only basic tasks to fully autonomous systems which are capable of going to sea with little or no communication and performing a task. With basic levels of control technology, a vehicle can be an effective observational platform. To use underwater vehicles for manipulative and experimental functions, the vehicle and its instrumentation need to have a more advanced level of control. Advanced vehicles that are capable of autonomous or semiautonomous functions need a fairly sophisticated controller because they need to be able to sense and respond to their environment. If properly applied, higher levels of automation and control in vehicles are useful because they increase the safety and effectiveness of the vehicle, allow the vehicle to function in ways that would not be possible if it were not automated, and provide a more cost-effective solution to accomplishing a task.

Assessment of Russia and Ukraine

There is little automation on the vehicle systems that the WTEC team observed in Russia and Ukraine. The systems tended to be manually controlled and operated. This is different from what you see in the United States, where the systems tend to be more and more autonomous in operation. Many of the vehicles in Russia and Ukraine (both manned and remotely operated) have no computer systems onboard. The manual systems in these vehicles seem to be reliable and to meet their specified objectives. In Western countries, automated control systems have reached the point where, for many applications, they are much more reliable than manual control systems.

An example of what can be considered an advanced vehicle automation system in the FSU is the Rus vehicle (Figure 8.3). This vehicle has an autopilot that receives input from a navigation system and controls the actuators to automatically fly through waypoints (see Malachite site report - Appendix B).

Another vehicle that has some automated systems is an autonomous unmanned vehicle, the MTS88, which was designed in Vladivostok. This vehicle has the capability to dive, execute a survey, and return under automated control.

Some of the potential causes for the lack of automation in Russia and Ukraine are lack of access to computers, the reliability of the electronics, and the relative cost of labor versus capital there.

Facility Automation

Most of the facilities that the team observed in Russia and Ukraine have almost no automation. The facility that seemed the most automated was the Krylov Shipbuilding Research Institute (see Krylov Shipbuilding Research Institute site report - Appendix B). Figure 8.4 is a picture of one of the pressure test facilities at Krylov. The pressure chamber is 3.2 m in diameter and can test pressure vessels at up to 1,000 atmospheres. This pressure test facility can automatically control the test through multiple pressure cycles, and can monitor the deformation of the test component in real time. The separate structural fatigue test facilities at Krylov automatically sequence 96 channels and measure 3,000 data points.

Computers and computer-aided-design software tools were identified by several organizations in Russia and Ukraine as the most important contribution that would make them more effective. Rauma Oceanics in Finland showed the company's extensive computer-aided-design and simulation facilities to the team. That company's scientists expressed the view that this capability is one of the major strengths that they bring to the underwater vehicle field.


Figure 8.3. An Autonomous Manned Submersible


Figure 8.4. Krylov Pressure Test Facility

Control From Local Sensors

Vehicle automation is a significant focus in the French research laboratories. One program that was described was the automation of oil well bore hole reentry (see INRIA site report). In this project, they try to use an automatically controlled vehicle to center and insert a tool in a bore hole using locally sensed data. The French also have projects in autonomous planning systems and the use of local sensors in the decision and planning process. A new French ROV is being designed that will have a fairly high level of integration and automation. The test vehicle VORTEX is being used to develop and test these new technologies (see IFREMER site report).

Manipulator Control

At Heriot-Watt University, a program is underway to develop higher levels of control for manipulator arms (see Heriot-Watt University site report). Scientists there are working to allow multiple arms to operate in the same workspace without collisions. This involves multisensor fusion and autonomous decision making within the robot control system itself. They have a demonstration of two robot arms moving in the same workspace. In the future there is the prospect of intelligent robots able to autonomously visualize, plan, and control manipulative tasks with a minimum of operator involvement.

At the Bauman Institute, scientists are working on manipulators and manipulator control. They are working on high positional accuracy (0.1 mm) manipulator control systems (see Bauman site report). Their control systems are consistent with other control systems that the panel saw in Russia and Ukraine that use joystick input to control either rate or force on a manipulator axis.

Software Architecture

In the Western European programs there is a significant focus on architectures for control of underwater vehicles. The goal of these programs seems to be to find methodologies that provide structures for data transfer within the vehicle system, and that provide clean, maintainable software that can function at increasingly higher levels of autonomy (see INRIA site report).

At INRIA, scientists are working on automatic generation of software for vehicle control (see INRIA site report). They have an automated robot design system to let them design the mechanical configuration of a robotic system using computer-aided-design tools. They then input the configuration from the CAD to an automated software design system. The system is configured by specifying modules, such as data sources and control filters. These modules are then connected together by data paths. C code is then generated that will run in a specified hardware and software environment. The code is simulated line-by-line and the performance verified.

Autonomous Vehicles

MAST II is a European Community program including two projects relating to AUVs. One of these, entitled "Advanced Systems Research for Unmanned Autonomous Vehicles," is being coordinated by Deacon Laboratory in the United Kingdom. Deacon Laboratory is also coordinating Autosub, a U.K.-funded program aimed at developing vehicles that can autonomously make transoceanic crossings and gather data and samples. This program is pushing the technology in long distance navigation, as well as in automation and control (see Chapter 5 and the Deacon Laboratory site report in Appendix E).

Conclusions

The level of automation in the vehicle systems and in the development laboratories and test environments is much lower in Russia and Ukraine than in Western Europe or in the United States. The infrastructure to develop automation and advanced control techniques is 10 to 20 years behind that in the West. There has been little or no emphasis on development of automation systems. There is little confidence in automation and little understanding of the pros and cons of its application. The autonomous control technology has not yet made its way into the vehicle programs of Russia or Ukraine, either in space or under water. Control systems tend to be hard-wired direct links to human operators.

Most of the test facilities in Russia and Ukraine are operated manually. There are a few test facilities that have automated data acquisition and test sequencing. There is no evidence of any software automation or hardware design. There was a consistent request from the designers for computer-aided-design and simulation tools.

The level of automation in the vehicle systems and in the development laboratories in Western Europe is similar to that in the United States. France has put a specific focus on control from local sensors, and the United Kingdom has put a focus on very long-range autonomous vehicles and the technologies that would enable them.


Published: June 1994; WTEC Hyper-Librarian