OPERATIONS AND PLANNING:

A case study from Mondi Limited, South Africa

Duonne M Erasmus

Natural Resources Manager

Forests Division

Abstract

Plantation forestry is an intensive land management practice for short rotation exotic timber species grown under South African climatic conditions. For this reason, the management and utilisation of accurate natural resource and operational information is of utmost importance to the efficiency of the business. Furthermore, Mondi's forestry landholdings are geographically dispersed which increases the complexity of creating and updating the spatial database by means of smooth and efficient information flows.

The process of planning, establishing and maintaining an Integrated Forestry Database (IFDB) based on a combination of UNIX ArcInfo and desktop ArcView technologies is described. Details of the processes followed during the user requirements analysis, the system design and the customisation stages are provided. A comprehensive description of the functionality of the ArcView software which was customised specifically for efficient forestry management and reporting is given.

Finally, aspects of the implementation of the IFDB system across 5 regional sites and 60 remote sites are highlighted and the progress to date is reported on.

1. INTRODUCTION

Mondi Limited, with total assets exceeding US$ 1,5 billion, was founded in 1967 and ranks as one of the most integrated timber products enterprises in the world. Six operating divisions cover all aspects of production and processing associated with the timber industry, including forestry, sawmilling, the manufacture and distribution of timber products, the manufacture of pulp, paper and board, the collection, sorting and recycling of waste paper and paper merchanting.

Mondi employs 17 500 people and has a production capacity of 1,9 million tons per annum including processed pulp, paper and timber products. Mondi owns and manages 436 000 hectares of forest land which will ultimately yield 4,5 million tons of wood. These forests, which are geographically dispersed along the eastern parts of South Africa, all consist of commercially planted exotic tree species including several species of the genus Pinus, Eucalyptus and Acacia. In addition, Mondi also produces 350 000 tons of sugarcane annually.

The main objective of the Forests Division is to provide an increasing supply of high quality, low cost wood fibre to the Mondi processing operations through excellence in forestry research and practice. In addition, approximately 30% of Mondi's landholdings are managed for conservation purposes according to stringent environmental standards. In order to achieve all these objectives, Mondi recognised the need for accurate natural resource information and has been committed to the collection of such information since 1985.

There are dramatic variations in soils, climate, terrain and indigenous species of fauna and flora across Mondi's landholdings which have to be gathered , stored and manipulated for use in the planning and management of our landholdings. Furthermore, the accurate reporting of daily operations and their associated costs are considered central to the success of the business and needed to be monitored spatially. These information needs, based on many different spatial variables, has necessitated the use of GIS technology as a core business tool in the Forests Division and resulted in the conceptualisation, development and phased implementation of Mondi's Integrated Forestry Database (IFDB) system.

2. BACKGROUND TO THE IFDB

During the period 1988 to 1994, GIS technology was closely monitored by the Natural Resources Manager as a tool to maintain the spatial database for the Forests Division, however, it was simply not cost-justifiable on this aspect alone. Furthermore, many of the managers in the Division could not fully comprehend the concept of GIS and the differences between GIS and our existing CAD system. It was soon realised that a concerted effort had to be made to educate all levels of management regarding GIS and to provide them with clear, practical examples of the information, planning and management benefits that could be derived using this technology.

In July 1994 a special Divisional management meeting was called specifically to address the issue of GIS and was attended by all senior management in the Division. A GIS consultant was called upon to clearly describe the technical issues of GIS vs CAD and to explain the major information benefits that could be expected over the longer term. The technical experts in forestry planning, natural resources and forestry management also gave presentations regarding GIS in forestry and proposed a previously agreed plan of investigation, assessment and implementation with associated time frames and resource allocations. This broad plan contained the following:

Phase I (1 September 1994 to 31 March 1995)

# undertake a user requirements analysis across the entire Division which consists of 2 subdivisions and 5 forestry regions.

# prepare a draft GIS data standards and database requirements document

# prepare a draft GIS system specifications document

# prepare a broad cost-benefit justification

# develop a broad IFDB development and implementation strategy for the Division with an associated budget and resource requirement plan

# select an appropriate pilot project area in each of the two subdivisions

Phase II (1 April 1995 to 31 August 1995)

# implement the GIS system at one site in each of the two subdivisions

# appoint new staff with appropriate GIS skills

# implement established operating standards

# undertake data conversion from CAD to GIS for pilot project areas

# evaluate the results of the pilot project

# establish an accurate budget for full implementation of GIS in the Division

# identify and prioritise core business applications using the GIS spatial database

# commence with the design and specifications of outstanding databases

Phase III (1 September 1995 to 31 July 1997)

# complete the customisation of desktop GIS

# complete the full implementation of GIS in 5 forestry regions containing 60 remote plantation sites

# complete the development of outstanding databases

# conduct data gathering exercises for populating outstanding databases

On 28 August 1994 the Divisional Executive Committee approved the commencement of Phase I of the IFDB project and its associated proposed management structure which consisted of the following:

Project Steering Committee: consisting of the Group IS Manager, the two Subdivisional General Managers, the Divisional Management Accountant and the IFDB Project Manager.

IFDB Project Management Team: consisting of the Project Manager, the Project Leader, the two Subdivisional IS Managers and the other members of the IFDB Project Team listed below.

IFDB Project Team: consisting of the Project Leader, two specialist Forestry Planners, the Divisional Forestry Systems Analyst and a specialist GIS Consultant.

This management structure worked very well because the Project Team were able to carry out their daily tasks whilst the Project Manager created the link between the Project Team and the Steering Committee. This ensured proper management, communication and ownership of the project at all levels of management whilst project deadlines were being met in record time.

The completed Phase I report was tabled before the Divisional Executive Committee on 19 April 1995 and permission was granted to commence with Phase II and Phase III immediately.

3. THE USER REQUIREMENTS ANALYSIS

The requirements questionnaire was designed to ensure that the respondents identified the types of information they require to do their current jobs and not simply to provide a "wish list". A schematic diagram showing the basic contents of the requirements questionnaire is shown in Figure 1. This enabled a data priority schedule to be compiled with confidence knowing that the data requirements were necessary to perform current tasks. Questionnaires were sent to all levels of forestry management, scientific personnel and selected administrative personnel.

Figure 1: Schematic diagram showing the contents of the questionnaire

The analysis also revealed that much of the information required by foresters and managers was not readily available to them, either because it simply did not exist or because it was only available at regional offices and not at remote sites. They also stressed that the accuracy of the available information was critical to the quality of their decision making.

Finally, the analysis revealed that there was very little ownership of information at remote sites due to a one-way information flow and that these sites felt that they were feeding an information "black hole". On further investigation it was found that there was a high level of data errors emanating from remote sites, a high level of general criticism of existing systems and complaints regarding database inaccuracies, the lack of timeous reports and much criticism of the general user-friendliness of the systems. Some of these complaints were entirely valid, however, in many instances they were purely symptoms of a lack of information ownership.

Other lessons that were learnt from this exercise were:

- all levels of management had problems in understanding and completing the questionnaire and the only accurate way of extracting the required information was by conducting personal interviews

- many managers do not understand what data really is and are not information orientated. They were often confused between data and tasks

- many managers had difficulty in describing their jobs in terms of the main objective of their job, the responsibilities of their job, the products and services they supply and the tasks that they undertake to provide the specified products or services

4. PLANNING THE IFDB

After carefully analysing the results of the user requirements questionnaire and combining these with the wealth of forestry, financial and technical expertise available within Mondi, the following key success factors were identified and used in planning the IFDB:

# the IFDB must provide timeous, up-to-date, accurate and easily accessible information to all levels of management and all departments using a combination of ArcInfo and ArcView software. A schematic diagram showing the role of the IFDB in the technical environment is shown in Figure 2 and shows some measure of the true complexity of commercial forestry

# ArcView stations at remote sites must contain a customised data input and editing module to enable spatial reporting of daily operations with associated costs of labour, stores and vehicles. This system would be called "the Foresters' Information Management System" (FIMS)

# FIMS must provide a user-friendly GUI that requires no more than 3 hours training for the forester level of management and must provide pre-defined standard reports to the forester on request

# operations reporting must be done on a daily basis using FIMS, with weekly updates to the regional ArcInfo systems. Weekly digital updates to the corporate database must adequately replace monthly written reports

# The turnaround times from FIMS sites to regional offices and back should be no longer than 3 days

# FIMS must facilitate all necessary validations that would provide warning and error messages for each transaction interactively. This would ensure that clean information would be entering the corporate databases at regional level with minimum high-level validations

# FIMS must provide a true "single-source" data capture facility that would update the map, financial and forest inventory databases simultaneously

Figure 2: The role of the IFDB in a technical forestry environment

5. IFDB SYSTEM DESIGN

The design of the system needed to adequately address the problem of geographically dispersed landholdings and management centres.

As already mentioned, the Forests Division consists of 2 subdivisions, 5 forestry regions and over 60 remote plantation sites called Working Plan Units (WPU). The two subdivisional offices also house one regional office each. The other three regional offices stand alone at different localities. A schematic diagram showing the data flows within the IFDB system between different localities is shown in Figure 3.

Figure 3: Schematic diagram showing the data flows within the IFDB system

5.1 Subdivisional Systems

Each subdivisional system consists of a UNIX ArcInfo workstation in an Ethernet network with 3 PC's, an A0 plotter and digitising tablet, a CD writer and a tape streamer.

This system caters for the information requirements of the subdivision by consolidating the spatial databases for all the regions of the respective subdivision as well as being used to compile and maintain the spatial database of the one resident region.

5.2 Regional Systems

The regional systems for the remaining three regions consist of a Pentium PC ARC/INFO station in an Ethernet network with two PC's, an A0 plotter and digitising tablet, a CD writer and a tape streamer.

These systems cater for the compilation and maintenance of the regional spatial databases and they update their respective subdivisional systems on a monthly basis as well as updating the remote WPU stations on a weekly basis.

5.3 WPU Systems

Each system consists of a single Pentium PC with a quad speed CD reader and an A4 ink-jet printer.

The system uses ArcView software and, using the customised FIMS system, caters for capturing of daily operations, labour, stores, vehicles and environmental sightings. It also facilitates the splitting, merging and renaming of forest compartments as well as the delineation of microclimatic zones and forestry areas that have been damaged. Standard ArcView functionality is also used to produce progress reports, work programmes and various maps.

6 CUSTOMISATION OF FIMS (Foresters' Information Management System)

Forestry, as a commercial operation, always uses a common unit of measure, namely; area (measured in hectares). This is true for measuring timber yield, income, costs, labour productivity, chemical application rates, thinning regimes and many others. For this reason it was apparent that GIS technology had the potential to become a core business tool in the Forests Division. It was also apparent that the spatial database had to be as accurate as possible from the start so that any operation that is undertaken in-field could be carefully monitored in terms of the hectares that the operation covered.

Forestry compartments include those polygons containing commercial timber species as well as those classified as "open areas" which are managed purely for conservation purposes. These compartments are the most basic management units in forestry and carry information of their management regimes, yields and costs.

After some investigation it was found that by using customised desktop ArcView software combined with an ArcInfo spatial database it was possible to accurately record operations in relation to the forestry compartments within the spatial database environment. Once this concept was investigated further it became apparent the customised software could be extended in an integrated fashion beyond the pure GIS aspects to provide a true single source data capture facility at remote sites. This concept immediately resolved the ongoing forestry problem of reconciling hectares, productivity, costs, income and yields and provided a source document that was completely validated and reconciled at the outset. This shifted the concept of GIS in the minds of managers from a "nice-to-have" facility to an essential core business tool that would be utilised on a daily basis to manage the business. All the other benefits that GIS technology could offer the Forests Division beca! me per ipheral benefits and the cost-benefit justification was done on the core business functionality alone.

FIMS has been programmed using a combination of Avenue, Visual Basic and Lotus Approach. The Lotus Approach application is only a temporary measure used to develop an application rapidly for thorough testing at the WPU level. Once the users are entirely satisfied with the functionality, the application will be re-written in a proper development language. The fundamental concept of FIMS is that the GIS part of the system manages and controls all spatial issues such as areas, compartment numbers, the forest inventory database and map production and conducts all the logical forestry validations. It then passes selected spatial information to the Lotus Approach application which uses this information to monitor, validate and report on cost elements such as labour, stores and vehicles.

FIMS has a customised suite of function buttons in the ArcView environment and the operator simply follows a "point-and-click" procedure for executing the various functions of the system.

6.1 Recording Daily Operations using FIMS

FIMS allows remote sites to capture operations for parts of a compartment or the entire compartment, depending on the nature and speed of the operation. In this way it is able to track the progress of any one operation across a compartment and also ensures that the correct hectares are reported thereby ensuring that the total hectares of a compartment are never exceeded.

The following procedures are followed when recording each operation:

- the FIMS operator selects the compartment where an operation is to be undertaken by pointing to the compartment on the map or typing in the compartment number

- the compartment appears on-screen at 1:10 000 scale and the print icon executes an A4 print which is taken in-field by the operations supervisor

- the area of completed during operation for the day is drawn on the compartment map in-field by the supervisor

- the completed map details are captured by the FIMS operator and the system prompts the operator for the operational details

- FIMS then validates the spatial and alphanumeric details of the operation, providing warnings and error messages where necessary

- the spatial information regarding each operation is stored in a shapefile for easy updating of the ArcInfo spatial database at the regional office using the shape-to-arc command

6.2 Compartment Management using FIMS

FIMS enables the remote site to manage their compartments in terms of splitting, merging and renaming compartments as well as being able to edit details of the compartment in the compartment inventory database if errors are encountered. For each of these functions there is a customised icon on the toolbar and spatial information is again stored in the shapefile for easy updating of the regional spatial database.

6.3 Recording Damaged Areas using FIMS

Damage to the forest can be caused by several agents such as mechanical, chemical, weather or pests and diseases, however, the most common agent is fire. FIMS enables the forester to draw an estimated polygon on the map where damage has occurred and prompts for the type of damage and its severity. Again these polygons are stored in a shapefile for updating the regional database. Information of this nature also provides a useful history of damage on the forest landholdings for future reference and planning.

6.4 Capturing Local Knowledge using FIMS

Local knowledge, in this instance, refers to microclimate, species of fauna and flora and areas of special interest (ASI). ASI includes features such as waterfalls, rock paintings, archaeological sites, historical sites and the like.

In the case of microclimate, the forester again simply draws a polygon to indicate a microclimate zone and the system prompts for the details climate type (e.g., hail, snow, mist, frost or severe wind exposure) and the severity of the microclimate. This information is also stored in a shapefile for updating the regional database. This type of local knowledge is invaluable to forestry planners especially if they are not familiar with the particular area which they are planning.

In the case of species and ASI, the recording of points rather than polygons is undertaken. This function is extremely simple. The forester selects the icon and then simply points to the position on the map where the sighting took place. The system then prompts for the required information regarding each sighting. The spatial information is again stored in a shapefile for updating the regional database and the alphanumeric information is used to update the corporate Environmental Conservation database.

6.5 Importing and Exporting data using FIMS

Importing and exporting are one button operations using FIMS. Exported data consists of the shapefiles, tables and database files that are used by the system and are needed to update the regional databases. Exported data is sent to the regional office via modem. Import files, which are far larger in size because they contain the entire updated spatial database for the WPU, are received by the WPU on CD. This spatial database incorporates all the latest information that was captured by their FIMS system during the previous week.

7. IFDB IMPLEMENTATION ISSUES

The geographically dispersed nature of the Forests Division makes the implementation of a system as sophisticated as the IFDB extremely complicated. The multiplier effects of minor problems at 60 remote sites could stretch any support systems beyond their limits in a very short time. Careful planning is required with attention to detail being of paramount importance. The critical implementation issues are as follows:

# preparing a detailed implementation plan with resources and timings carefully specified

# developing the spatial database in a phased manner that precedes the system roll-out

# ensuring that all spatial database standards have been rigidly adhered to

# ensuring that all hardware requirements have been carefully planned and that identical equipment is purchased so that a "plug-in, plug-out" support strategy can be implemented

# ensuring that the ordering of hardware and software is correctly phased within the roll-out program

# using a resource requirement plan, carefully select an implementation team who have the required forestry experience and are computer orientated

# developing an appropriate training schedule for the implementation team

# planning the phased roll-out of the system to remote sites ensuring that sufficient "bedding-down" time is allowed for each remote site whilst hand-held support is available

# ensuring that remote sites have a rigid updating routine to reduce updating bottlenecks at regional offices and to meet financial reporting deadlines