WHITE PAPER

There exists within the disaster management community an on-going need for real-time data and information, especially during catastrophic events. Currently, much of the data is collected by twin engine or jet aircraft with limited altitude and duration capabilities. Flight safety is also an issue. Clearly, much of the needed data could be delivered via over-the-horizon transfer through a UAV platform to mission managers at various locations on the ground. In fact, because of the ability to stay aloft for long periods of time, and to fly above dangerous situations, UAVs are ideally suited for disaster missions.

There are numerous situations that can be considered disastrous for the human population. Some, such as fire or flood, can continue over a period of days. Disaster management officials rely on data from the site to respond in an optimum way with warnings, evacuations, rescue, relief, and to the extent possible, damage control. Although different types of disasters call for different types of response, most situations can be improved by having visual images and other remotely sensed data available.

"Disaster Management" is actually made up of a number of activities, including:

Locating a UAV with the right capabilities at a disaster site affords an opportunity to impact disaster outcomes because it can provide to the right people the right information at the right time regarding the situation on the ground. To achieve this capability on an operational basis requires an ability to put a suitable platform with appropriate sensors in place above a disaster location and provide real-time, over-the-horizon data transfer to personnel who can put the information to use in a timely and helpful way. Figure 1 shows the demonstration elements.

With all these potential benefits, it is desirable to demonstrate to user agencies the ability to perform disaster management missions as described. The following demonstration project is the first in a program designed to prove the feasibility of supporting disaster missions with UAV technology and suitable communications packages on-board.. A several-year program is envisioned, in which a broad range of disaster-related activities are demonstrated to the appropriate user communities.

GOALS OF THE DEMONSTRATION PROJECT

The goals of carrying out disaster management demonstration missions are:

The objective of the demonstration missions is to successfully integrate the following capabilities into a system and operate in a disaster management scenario:

There are many potential missions. These are listed below:

Table 1. Mission Parameters
 

TECHNOLOGY STATUS DISCUSSION

UAV Status and Opportunities

NASA has access to a number of UAVs that are available or becoming available for disaster demonstration missions. Table 2 below lists the status of several platforms.

Table 2 UAV Platforms
 

O-T-H Status and Opportunities

Over-the-Horizon data transfer is a major goal of this demonstration program. At present there is very limited commercial service available for OTH applications. Previous NASA experiments have used TDRSS, a NASA satellite normally restricted to special use. Military systems have been used in some experiments that had joint military sponsorship. The available commercial systems are listed in Table 3 below, along with their availability for UAV demonstrations. For early demonstrations, it is likely that some compromise will be needed to achieve some degree of OTH communication.

Table 3 Commercial Over-the-Horizon (OTH) Systems
 

Sensor Status and Opportunities

Although many sensor payloads are available, in principle, for disaster management demonstration missions, in reality only a few are supportable by the current UAV community. Weight, power and size are issues; integration is required to make use of limited space; many payload areas have little or no environmental control; and payloads must be controlled autonomously for UAV applications. Table 4 below lists several disaster management payloads and their characteristics.

Table 4 Sensor Payload Characteristics
 

Real-time telemetry and analysis: Status and Opportunities

ERAST demonstration missions provide the opportunity to address several challenges that NASA is facing as the quantity of data produced with new sensor technology and the scientific data requirements from space or airborne missions tax the bandwidth of present telemetry downlink systems. The basic physics of transmission favor the investment of electrical power assets not in higher bandwidth transmission but in intensive low power onboard data processing, where Moore's law promises ever improving computational capability. The development of "smart sensors" will reduce the volume of the downlinked data stream.

In the near term, several strategies may be exploited. A starting point is the development of algorithms that select only data that is relevant to very specific scientific objectives. This may be as simple as selecting specific targets or not transmitting images from clouded areas. Another technique that is common with image data is compression, which may be categorized more generally as the use of mathematical tools to compress information into a minimum lossless (or prescribed lossy) volume as limited by information theory. Often remotely sensed image data is processed with sophisticated classification software to identify the spatial domain of a specific phenomenon. As an example, a huge multi- or hyperspectral data set might be reduced to a single binary image mapping a particular vegetation type. At a more complex level, algorithms may be developed to select elements from a large data set based on optimal scientific content as measured by a field programmable merit function.

For the future, it is possible to envision the use of neural networks or rule-based artificial intelligence tools to maximize the acquisition of useful data, particularly if autonomous direction of the aircraft flight path is incorporated into a full robotic mission concept. This concept is described below as real-time retasking.

Real-time data distribution: Status and Opportunities

Real time data distribution is currently planned to be handled in three different ways, depending on the tools and internet connectivity available. Scenarios 1 and 2 could reasonably be carried out today. Scenario 3 is in development and should be available within the next year or so, depending on funding profiles.

Scenario 1, Field Access:

A field commander would access the Internet server with a standard web browser, and download the most recent incident image or relevant data, or any historical (archived) images or data.

Scenario 2, Command Access:

A regional commander in the command center would have access to either processed or raw image or instrument data, and, additionally, ancillary data sets, available in different formats, on the Internet

Scenario 3, Digital Earth Augmented Access:

A field commander with a reasonably high speed Internet access, a regional commander in the command center, and (possibly) distributed disaster center personnel could, using freeware tools being co-developed by NASA and the OpenGIS Consortium, utilize resources on the "Digital Earth" in concert with real time imagery or other data. In this scenario, a user could, for example, overlay a road layer from one "Digital Earth" server over the real time incident assessment image, which was, itself, overlaid on a USGS digital orthophoto that was retrieved and re-projected by another server. At any time, current asset locations and descriptions can be downloaded and displayed.

Real-time data availability makes possible real-time retasking, as described below.

Real-time system retasking: Status and Opportunities

The most important goal of providing real-time data to users on the ground is to support disaster response by allowing real-time decision making regarding flight coordinates and data acquisition functions of the UAV and its payload. The concept is to allow observers at a mission control station to task or retask the UAV and its payload by directing it while in flight to a specific location for a desired measurement or observation. Some examples:

Table 5 Real-time Data Demonstrations
 

PROCESS OF SELECTING DEMONSTRATION MISSIONS

The goal of UAV disaster management demonstrations is to show the feasibility not only of the technological components, but the integration of the UAV system and interface with likely users. A general process for selecting missions is described below. The first demo, to be carried out in 2000, is also described briefly below. Details of the 2000 mission, prepared by Steve Dunagan of NASA-Ames, are appended to this white paper.

General selection process

A number of attributes of the desired disaster management mission and the available technology must be considered when preparing for a demonstration. The key element is to define an exercise that can succeed, both technologically, and in serving the needs of the user community.

Some items to consider:

Having created a matrix of possible missions, the items listed above as selection criteria were considered. Primarily on the basis of limited time and budget, and previous experience with the fire management community, it was decided to proceed with the following disaster management demonstration:

DEMONSTRATION MISSION MANAGEMENT

This section describes the process under which demonstrations will be carried out. The FY2000 mission is used as an example.

Overview

One goal of the ERAST sensor element, managed by NASA Ames, is to integrate UAV and sensor technology with the intention to provide an extended UAV user base and to identify new regulatory and technology needs. Limited resources require highly leveraged demonstrations. Disaster management missions have been identified as being very well suited to UAV operations and hence are highlighted.

Demo mission objective

FY2000 and 2001 UAV demonstrations will utilize existing platforms, sensors, and analysis tools to provide relevant near-real-time disaster imagery products to personnel in the field. Multiple missions are envisioned. If appropriate, a quick-response capability will be demonstrated.

Demo mission approach

Altus will be equipped with the AIRDAS system (utilized by fire researchers) and standard communications connections for data download. Imagery combined with various other parameter relevant to the demo mission, such as GIS products, will be provided to disaster professionals. Opportunities will be explored to support real disasters.

Resources and Personnel

Project manager — Steve Wegener

UAV Flight Operations — General Atomics Aeronautical Systems

Payload development and integration — Steve Dunagan

Real-time data products, communications, Internet — Don Sullivan

Fire disaster incident response simulation — U.S. Forest Service

Demo mission schedule

In download.

Demo mission budget (follows final section)

LONG-TERM PLANS

The long term plans for UAV disaster mission demonstration include:

Along with future disaster missions, improved technology and system operation are anticipated. Some identified future technology and system development needs are listed in Table 6 below.

Table 6 Summary of identified future technology / system development needs
 

FY2000 ERAST Disaster Management Demonstration:

Fire Management Mission

System Requirements Document: Rev B 12-2-99