Early Times of Digitizing Biostratigraphy: ADIE, SAPIENS and the Associated Expert Systems

Between 1986 and 2001, the Biostratigraphy Unit of Maraven, S.A., and subsequently of Petróleos de Venezuela, S.A. (PDVSA), developed a suite of computer systems to automate and digitize the analysis and interpretation of stratigraphic samples. The first of these was ADIE (Análisis Digital de Imágenes Estratigráficas), a digital image analysis system for the quantitative morphological characterization of particulate organic matter, petrographic thin sections and polished rock fragments, patented in the United States in 1990 (US 4,918,739), with patents also granted in Australia and the United Kingdom. ADIE evolved through three hardware generations. The second generation incorporated, in 1992–1993, an expert system for the classification of palynological organic matter under transmitted white light and incident white and fluorescent light. Its third generation introduced phase analysis based on luminance thresholds and mean shift color segmentation. A second expert system, ESTELA, supported the identification of key pollen and spore markers for northern South America. This work culminated in SAPIENS (Sistema Integrado de Aplicaciones para Estratigrafía), an integrated platform that linked species counting, lithological description, interpretation, image analysis and the generation of distribution charts to a corporate stratigraphic database connected to FINDER, the PDVSA exploration and production database. This article documents the objectives, architecture, methods and published applications of these systems, assesses them against the technology and common practice of the period and preserves a record of an early effort to digitize biostratigraphic practice in the petroleum industry.