WHY, HOW, AND THEN WHAT?

Daniel Elroi
GIS Manager
Knight Piésold LLC
1050 Seventeenth Street, Suite 500
Denver, Colorado 80265

Mike Price
Mining Industry Manager
ESRI
380 New York Street
Redlands, California 92373

ABSTRACT

This paper addresses the sharing of 2D and 3D data between GIS and mine planning software. Two-dimensional (2D) data, and now three-dimensional (3D) data as well, can be formatted such that both types of software can use, exchange, and display the same datasets. This increases the ability to share data between such groups as land, environment, exploration, and mine planning. It overcomes shortcomings of mine planning software in areas of open database connectivity, spatial analysis, and data input and output. It also helps to overcome shortcomings of GIS software in the areas of geologic modeling, three-dimensional interpolation, and visualization.

INTRODUCTION

In order to discuss the reasons for interaction between mine planning and GIS software packages, the following general statements need to be made. GIS traditionally tends to view the world from above, either as flat surfaces or, at best, as discrete horizontally oriented surfaces such as topography. The term "2½D" is often used to describe the way in which GIS views natural phenomena. Therefore, GIS has been traditionally limited in the types of analysis it can perform with respect to geologic interpolations, block modeling, mine planning, and three-dimensional visualization. Mine planning software packages, which are traditionally strong in these areas, nonetheless tend to be weak in their ability to manage broad databases, to permit easy spatial and tabular data input and output, and to perform robust spatial analysis. Frequently there are also differences in approach to user interface techniques, openness to interaction with other software packages, and methods of customization. Since both types of software deal with spatial data it would appear that benefits could be realized from interfacing them with each other. Until recently this was only achieved with difficulty.

Several advances have occurred in the past two years that make it far easier to exchange data between GIS and mine planning software, and which should therefore be explored. From a technological standpoint, sharing data between these types of software offers a means for overcoming technical limitations in each. From an operational standpoint, data sharing enables professionals from such departments as exploration, mine planning, property management, hydrology, and environmental management, to share data and computing resources.

THE NEED TO EXCHANGE DATA

The potential advantages to being able to exchange spatial data between GIS and mine planning software need to be spelled out. From the standpoint of either GIS or mine planning software users, accurate data transfer can allow the following functions, among others.

Furthermore, the results of a modeling or design exercise in a mine planning package may be communicated through a GIS package, in combination with other typical GIS layers such as topography, infrastructure, and orthophotography. Conversely, data necessary for modeling in a mine planning package may first be organized and reconciled in a GIS package, incorporating data from multiple sources, scales, and projections. For these reasons, and for the simple reason that the popularity of both types of software is increasing, there is growing pressure to enable intelligent and comprehensive exchange of information between GIS and mine planning software.

BARRIERS AND LIMITATIONS

There are several factors which act as barriers and limitations to the ability to share and exchange data between GIS and mine planning software. It is important to understand these, in order to appreciate some of the recent developments in this area. Although some of the source data used in the two types of software are similar, namely surface contours or digital elevation models, sampling points, and other surface vector data, many other types of data are quite different. The prime examples, of course, are data collected from boreholes, since these are inherently three-dimensional. Moreover, the way in which data are structured and stored in these two types of systems is either very different, or subtly but sufficiently different as to render them incompatible. For example, a block model structure often used in mine planning is very different from the vector data structure frequently used in GIS, but also quite different from the discrete-layer grid structure used by many of today’s GIS packages. In the vector data arena, GIS makes use of topology to display and analyze polygons using single, non-duplicated, boundary lines. Many mine planning packages, which are designed to read CADD files rather than GIS files, must process each polygon as a unique combination of closed lines in order to analyze or display them as closed polygons. Finally, whereas GIS packages can often create and read TIN structures, the use of TIN in mine planning software to describe solids rather than mere surfaces, far exceeds the capabilities of GIS packages to create or understand such complex TINs.

The vendors of GIS and mine planning software are not blind to users’ desire to exchange data with other software. However, the types of standards being developed for one industry, as in the example of the Spatial Data Transfer Standards (SDTS) being developed in the GIS industry, do not necessarily impact the other type of software industry. This is mostly due to the demands placed on these vendors by their users and the requirements specific to their applications. This has resulted in a limited number of tools available to the user who is interested in interfacing these software packages with each other.

TRADITIONALLY AVAILABLE EXCHANGE TOOLS

A set of tools has been traditionally available to users who wished to undertake the sometimes arduous task of exchanging data between GIS and mine planning software. One method has been the exchange of source data in ASCII format. For example, DTMs and XYZ lists have been shared by such packages for some time. The disadvantages to this method are that there is only a limited amount of useful information that can be exchanged in this way, and that any analytical process applied to the source data by either type of software is lost in the transfer. If data are transferred by these means, and both software are in fact able to use the data, they are still likely to interpret the same data in different ways. For example, a DTM exchanged in ASCII format between GIS and mine planning software, will in all likelihood be interpreted differently by these packages, and therefore produce inconsistent results.

Another method for exchanging data between GIS and mine planning software has been the use of DXF files. Although the DXF format can adequately exchange basic point, line, and text information, it is essentially flawed when it comes to exchanging polygonal and three-dimensional information. For example, whereas the DXF format supports multi-elevation lines (such as streams and boreholes), many GIS packages have not traditionally been able to take advantage of this information. Likewise, polygonal data exchanged through DXF is often perceived as a series of disconnected lines by mine planning software, which then need to be reconnected all over again to form polygons.

The rich attribute data that can be attached to spatial entities in a GIS cannot be easily transferred to and from mine planning software packages using either of the methods mentioned above. For this reason, and the other reasons mentioned, these methods cannot be considered to be satisfactory. In fact they are only available incidentally, since they have been provided to handle other problems, such as the exchange of data with CADD packages and data collection devices. Because these methods are difficult to use, they have not contributed greatly to the regular exchange of information between the two types of software packages.

NEW CAPABILITIES AND NEW OPPORTUNITIES

Several advances in the past two years have opened the door to a closer integration between GIS and mine planning software technologies. These include:

These new capabilities in turn translate into new opportunities for data exchange. For example, the emergence of published data storage formats is making the writing of data translators a less difficult task, both technically and philosophically. With the ability to communicate more easily between software packages also comes the ability to perform conversions behind the scenes, without directly involving the end-user of the software. The ability of some GIS packages to understand three-dimensional data more comprehensively allows much of the translation between three- and two-dimensional data to occur within a certain package, in this case the GIS package, rather than at an external data translator level. Data translators that are external to both types of packages are more prone to errors. They can more easily fall out of phase with developments in either of the two commercial packages that they link.

A CASE STUDY

In order to determine the viability of creating a tighter link between GIS and mine planning software, two software packages were selected: ESRI’s ArcView 3 GIS package with the 3D Analyst extension and Mintec's MEDSYSTEM mine planning package. A sample dataset containing several dozen fictitious boreholes was developed in MEDSYSTEM, along with 3D TINs topography, and block models of the orebody. These were then exported as ASCII or DXF files, and imported into ArcView 3D Analyst. All objects imported correctly and appeared in three dimensions in the GIS package. Each of the boreholes came in as a series of segments, all of which were open to interactive query for their database attributes. 2D representations of the block model were imported into 3D Analyst, placed in vertical space by elevation, and extruded upward by block height. Ore grades and other attributed data were thematically mapped in the extruded model. Because the 3D module is part of the GIS package, each record could also be related to a variety of other databases, accessed directly, through an ODBC, or a relational database gateway. Because of the integration between the 3D and 2D viewers in this software, boreholes could be overlaid with other spatial layers such as property boundaries in either viewer. Also, through standard features of the GIS package it was possible to link scanned documents, such as borehole photographs and core logs, to the boreholes, and to retrieve them for display.

Although the surface TIN was not imported as a TIN, but rather as a collection of individual 3D polygons, it was easily converted back to a TIN in ArcView. This made it possible to drape a variety of layers on this TIN from the wide variety of vector data sources that a GIS package is usually able to read, such as SHP, VPF, DLG, DGN, DWG, and coverage files. This also made it possible to easily overlay and drape raster files in a variety of formats, such as BIL, LAN, GRID, IMG, TIF, and JPG. Finally, the three types of data which were imported from the mine planning software, and the layers which were added in the GIS, were conveniently exported from the GIS to a VRML file format which can be viewed in three dimensions on a Web browser.

FUTURE DIRECTIONS

As users of GIS and mine planning software influence software vendors to enable easier data exchange, transfer of data will increase. At some point in the future it may be that vendors will begin to implement common data exchange standards, such as the Spatial Data Transfer Standard mentioned before. Whether or not vendors take a proactive role, users or third party developers will begin to develop more efficient data exchange methods.

Concurrent with, or subsequent to, improved data exchange capabilities, there may come a time when GIS and mine planning software enable each other to utilize their respective analytical strengths in a distributed client/server fashion. If this occurs, either software may act upon an entire dataset, depending on specific analytical needs, and in a fashion that is transparent to the user. For example, a user may view the dataset through a Web browser, and pose a question to the dataset. The question will be directed to the software that can most appropriately answer it, and the answer returned to the user. The technology to implement such a scenario exists in the technology marketplace today, and is already implemented in a variety of fields. It remains to be seen, however, when these methods are adopted in the GIS and mine planning software industries.