Utilizing big drilling data requires an innovative approach. The service company’s drill bits business is largely based upon an in-house drilling record system (DRS) that captures global bit record performance data. The DRS contains over 1.8 million wells drilled worldwide since 1980 with nearly 5.4 million total BHA runs from over 100 countries. In the last 10 years alone, over 1.4 million bit runs drilling over 2.8 billion ft of formation have been recorded. To utilize this vast amount of data for drill bit performance evaluation, analysis, and monitoring, the innovative approach described in this paper was developed and implemented. Traditionally, the performance of a drill bit run–often measured in terms of drilled footage and ROP–has been evaluated versus similar offset runs. Offset runs are chosen in various ways, but are typically done manually by bit engineers, meaning that offset run selection is subjective based on personal experience and bias. Furthermore, people often only evaluate the performance of test bit designs. Instead, we wanted to analyze and monitor the performance of all drill bit runs. To alleviate these biases and enable a wider breadth of considered runs, an objective offset run selection workflow was developed and implemented within DRS. Offset runs are selected based on a sophisticated filtering and scoring routine that considers many characteristics such as geographic location, time, wellbore and drilling system design, along with lithology. As new data enters DRS continuously, this workflow runs on a regular basis using an automated pipeline. The performance evaluation results of the automated offset selection workflow are available to all data analysts (engineers and salespeople) both inside DRS and extensible applications to aid in performance monitoring and new product development target-setting. Product performance is now objectively evaluated at-scale across geographies and always utilizing apples-to-apples comparisons. The workflow has proven itself quite useful and delivered business value already but also exemplifies the need for both enhanced data quality and improved bit record data capture rate. These are ongoing efforts to further enhance and improve this workflow. Automated workflows like this one can help our industry by eliminating repetitive biased tasks and allowing people to focus on more creative processes leveraging objective data. Developing new drill bit designs, material selections, or component selections to overcome new challenges are creative processes which contribute to increased drilling performance and lower costs for the industry.

The energy industry is undergoing a digital transformation, whose goals include increased operational efficiency and reduced energy extraction costs. Data science and machine learning (ML) are enabling the drilling engineering community to contribute to the success of these goals. An ML-based digital solution has been developed to assist the drilling engineer select an optimum bottomhole assembly (BHA) and drilling fluid technology during the well design phase. Traditionally, these selections depended on offset well analysis, which is a manual and time-consuming undertaking. As an alternative, the new digital solution, launched in the form of a web app, automatically selects similar offset wells, and evaluates the available BHA and drilling fluid options from those similar wells. The web app displays these options to the drilling engineer, who is now empowered to make fully informed data-driven decisions. To power the new digital solution, an extensive effort was made to gather, clean, and prepare global operational data into a new database. This operational database includes the selection decisions and performance results of drill bits, motor power sections, rotary steerable systems, BHA configurations, and drilling fluids. After the drilling engineer defines the parameters of the planned drilling run, a multidimensional distance-based approach is used to automatically select the most similar previous drilling runs within the context of the technology selection. The drilling engineer can also fine tune the offset selection based on experience using filters in the web app. Once the most similar offset runs are determined, the technology selection decisions are scored for numerous key performance indicators (KPIs). These KPIs, along with user-defined weights, drive the overall scores. Finally, technology selection recommendations are based on the overall scores and other contextual data such as local availability and cost. The new digital solution has been deployed to a global group of drilling engineers. Feedback sessions are held regularly, and the development team uses this feedback to rapidly iterate and improve user experience. While today's drilling engineers have access to a vast amount of data and information, it often cannot be used in a practical and efficient way. The new solution places all previous drilling system technology selection choices and results into the hands of the drilling engineers, allowing them to make their best decisions. This approach demonstrates how ML and innovative software deployment methods can truly assist the human decision-making process and succeed in accomplishing the goals of digital transformation. To our knowledge, this is a unique approach to drilling system design optimization. Not only is the approach unique, but the database developed as a portion of this effort is likely the largest drilling operations database within the industry. This paper presents all phases of the project, including the details of database creation, data preparation, development of the ML models, and the creation and iteration of the user interface. Finally, this paper presents the future of this effort as part of the company's vision to be our customers’ performance partner of choice.