RASTRO: INTERNET BASED TRACKING SYSTEM FOR FISHERIES CONTROL

Rodrigo Becke Cabral, Rafael Medeiros Sperb, Rodrigo Zanato Tripodi, Roberto Wahrlich, Jesiel de Souza

Universidade do Vale do Itajaí (BR)

Introduction
Since 1999 the offshore fishery industry in Brazil has invested expressive amounts of resources in searching for alternative deep-water commercial stocks, such as monkfish and red crab. However, the industry fleet has no tradition or either is prepared to fish in deep-waters. Hence, the Federal Government created a control program permitting rented vessels from Spain, United Kingdom, Portugal, Germany, Russia, Japan and South Korea to operate along the Brazilian coast. In order to carry out this operation, all vessels must have onboard tracking equipment, an independent observer and is licensed to operate only between 200 and 600 meters depth.
This paper presents how the tracking equipment works in compliance with RASTRO, an Internet based tracking system that was planned, developed and implanted to monitor fishing grounds (CABRAL, SPERB and TRIPODI, 2002). The result is a successful balance between the commercial need of the fishing industry and the responsibility of the government in controlling fishing resources.

Fishery Management
Fishery management can be defined as a set of governmental policies that aim to develop and maintain fishing activities in the long-term, making considerations upon fishery stock replenishment capacity as well as the economical, social, political and strategic needs of a nation.
Due to its inherent free access nature, commercial fishing not bound by adequate management actions will lead to the inevitable depletion of fishery stocks. An efficient fishery management policy reaches its goal when the end result is the best possible economical and social benefit, restricted by an optimal use of fishery resources.
Since the 70´s Brazil has adopted many actions in fishery management focused in the most profitable fishing stocks, but with no means to ensure the sustainability of such resources, and therefore harming the long-term growth of the national fishery industry. The largest difficulty has always been in the lack of human resources and adequate infrastructure to control and monitor the fishery activity.
In time, technology fostered the concept of tracking vessels with satellites, with a minimum requirement of resources. This idea allowed the development of an important tracking program to control and monitor the fishery industry, aiming to increase the efficiency of fishery management in Brazil (MMA, Brasil 1997).
First, rented foreign vessels have been obliged to install equipments and to provide geographical information on their activities. In the near future, the Brazilian government considers expanding fishery vessel tracking to the national fleet, as initial results confirm the low cost and feasibility of the program.

The Beginning
In 2000, the government started the program establishing poor definitions on the tracking technology and on the control practices and goals for the program. The group for fishery studies (GEP) from the CTTMar College at Univali (Brazil) was assigned to receive, plot and report on vessel activities listed for the program. As it should not force the industry to bear with one single provider for tracking vessels, the government put that "delivering date, latitude and longitude to GEP" certifies any fishery company to legally operate in the Brazilian shore.
On the other hand, the fishery industry was just preoccupied in hiring a low cost service to meet government requirements and to legally run their operations. Therefore, most fishery companies chose a tracking system to provide a simple, low cost service, which consisted in delivering to GEP emails listing vessel name, date, latitude and longitude.
In early 2001, GEP found itself buried deep in a number of data constantly arriving by email, and with a complex procedure involving one full-time human resource to receive, read, and type in a GIS software incoming position data. That is not to mention bi-weekly reports to write and the possibility of human-failure due to non-automated steps.
At that time, researchers from a neighboring Applied Computing Lab (G10) were experimenting with webgis technologies, and found the problem that GEP was undergoing a good challenge to practice their knowledge.
Now, with more than two years of usage and experience, RASTRO has been conceived as a fully automated Internet based tracking system available to GEP - for monitoring purposes; the government - for auditing; and the fishery industry - for strategic ops.
RASTRO
The RASTRO system is a collection of web and shell scripts that perform a variety of tasks including data fetching and processing, shape file generation, dynamic map generation of maps for the web, and on demand reports.
Currently, several companies have certified equipment to provide tracking positions in compliance with RASTRO. The system collects vessels´ GPS data and displays maps and reports on demand showing vessels engaged in fishery activity.

System Architecture
Brazil is a continental country. Using the Internet was an appropriate choice to reach all potential users of RASTRO. As such, the technologies used in the system were selected using two major criteria: web born capabilities and open source nature. Exceptionally, the database management system (DBMS) adopted was Oracle, with the Spatial Database Objects expansion cartridge (SDO). Underlying the DBMS is the Red Hat Linux operating system, integrated with the Apache web server, the PHP script-processing module and the MapServer technology sponsored by University of Minnesota and the TerraSIP project. Combined, this is the platform that sustains the RASTRO system, which uses PHP scripts for web user interfaces; and Linux shell scripts for automation as noted in Figure 1.

Figure 1. System Architecture

The RASTRO system is structured in two modules: data reception and web map navigation.
Data Reception
Data reception is totally automated. This module is the interface to external tracking information providers that continuously collects GPS positions from off shore vessels. Hence, at this point is where data is delivered to the RASTRO system. In the most usual communication topology, all certified third-part tracking equipments are in contact with inland servers, which in turn supply formatted vessel position information to the data reception module. This module processes data and stores position information in the Oracle GIS enabled database.
The original challenge was to read emails that contained GPS positions. Using a popular tool found in Linux distributions - the fetchmail, a shell script was developed to retrieve and parse emails from an email server using the POP3 protocol. It was the automated version of "reading an email and charting the incoming vessel position from the message". The script is scheduled to run in regular periods, providing close to real time access to information via web.
Currently, many other forms of data reception and/or retrieval are functional. Legacy standards have been studied and implemented to further certify third-part equipments, including access to GPS data by telnet, and so on. In fact, the lack of standardization raises an important issue: the difficult to expand and evolve the RASTRO system as data reception becomes more complex. This problem is unsolved and currently undergoing revision.

Web Map Navigation
Web map navigation is an interactive web user interface that displays layers of last positions (Figure 2) and trace data (Figure 3) on a base map. Interface functions are available according to the user profile and his or her permissions.
RASTRO accepts users to sign in with one of the following profiles:

  1. Auditor: can access all information and perform spatial queries from any time frame (e.g., see sample time period defined in Figure 3). Further, auditors may generate a report in PDF format for officially documenting fishery activity as defined by Brazilian law (e.g., see the "printer button" in Figure 3);
  2. Observer: can access all information, but can only access data from the last ten days. This information helps inland team coordinators to mainly follow vessels´ departing and arriving points.
  3. Company: can access information restricted to vessels owned by the given company, but it is not bound to perform spatial queries from any time frame. Web availability of data helps company employees to coordinate operations from anywhere, anytime.

The last positions view resumes the distribution of all vessels along the Brazilian coast. It displays in black with a crosshair mark the names of vessels that have been successfully located within the last 4 hours. On the other hand, vessels out of reach from more than 4 hours are highlighted in red with an x-mark, showing the last known position.

Figure 2. Navigating in web maps - last positions

The trace data map view is used to monitor fishery activity, showing all received positions within a given time frame. Each position mark is connected to the following position with a black solid line unless a not-found signal has been received between two position signals - in this case, a red dotted line will be plotted. If possible, alongside each position mark is a number to identify the position data from a table of positions below the map.

Figure 3. Trace path for a single vessel

When a vessel is engaged in fishery activity, speed estimates between points are lower than four knots. A special "fish button" (Figure 3) allows the user to observe only low speed activity; clearly marking the space that probably the vessel was active (Figure 4). This information is essential when writing official reports to government to confirm that the operation is running according to fishery management policies.

Figure 4. Fishery activity spots

Other features commonly found in webgis are also present in RASTRO, like zoom in, zoom out, reset view, and query data.

User Feedback
When RASTRO first entered in operation, immediate results followed as the GEP group, for the first time, was able to work with up-to-date data, no longer having to issue reports a month or two behind schedule.
Later on, the system included a view to inland team coordinators who could use the latest position information to instruct onboard observers the procedures upon arrival, including preparations for reception of material collected during the journey.
Finally, companies from the fishery industry required access to RASTRO. The visual presentation of GPS data can now be studied in combination to fishing spots to lay down logistics and strategic operations.

Why Open Source?
It is quite interesting that the system has been developed using open source technologies (Raymond, 2001). It is understated that open source software (OSS) or free software (FS) guarantee free use of all digital artifacts of a given system, for any purpose, including the modification and redistribution of the original code (Wheeler, 2002).
OSS/FS success is in the transparency of experimentation for adoption to everyone's particular needs (Shirky, 2002). RASTRO is built for Linux, Apache and PHP -well-known names in the open source community.
However, we would like to make a special regard for MapServer. The open source MapServer technology has found fast use as a web CGI tool, as many websites have been easily putting online maps like the ones listed in MapServer's official website gallery. But its potential is really met when advanced script processing like PHP is integrated to the technology - that is the MapScript PHP extension.
RASTRO has been developed using MapScript. All maps are plotted on demand. Information cannot be accessed directly, according to implemented security guidelines. The trace path in Figure 3 is plotted as a dynamic shape created from the database data. MapScript is quite fast, and the user can observe no delay during map drawing.

Future Work
Considering the multitude of commercial tracking systems available in the worldwide market, RASTRO presents itself more as a tracking center, providing position data in several levels for users, and building a strong appeal for integration of numerous tracking systems as it is currently functioning today.
The idea of integration now leads the G10 research group to standardize tracking data capture or delivery, including not only position information, but also further measurements like temperature, salinity, depth, and others. Undergoing work is focused in setting a long lasting standard to aid the development of fishery vessel-tracking industry in Brazil. The first draft of the standard is expected to be ready by the end of 2003.
The standardization on this kind of telemetry data will lay a milestone on the efforts of the G10 lab and GEP. In fact, it can be seen as a nation-wide consensus on how to report fishery activity data to support fishery management - a standard soon to be officially recognized by law, and that may further found other applications in Brazil.

References