Cognitive Geology
Our ground-breaking multi-million dollar AI product was designed to revolutionize how international Oil & Gas companies classify geological formations while preserving subject matter expertise. With our solution, we expedite the process of subsurface modeling, increase accuracy and consistency, and generate deterministic data to make more effective decisions. Say goodbye to manual data processing and welcome the power of AI to your business with our innovative solution!
Read here to see IBM's publishing of my product.
Geology subject matter experts face a challenge in preserving and sharing their knowledge with junior geologists, while also being able to focus on their work. There is a need for a product that can enable the experts to store their knowledge in a scalable and accessible format that allows for easy sharing with junior geologists. The product should offer a solution that can reduce interruptions and help junior geologists find the answers they need quickly and efficiently while freeing up time for subject matter experts to focus on their work.
Problem Statement
Users
Name: Alex Johnson
Age: 25
Occupation: Junior Geologist
Education: Bachelor's degree in Geology
Goals: Alex wants to gain more field experience and work on larger projects. He is also interested in taking on more responsibilities and advancing his career in the next few years. He wants to stay informed on new technologies and advancements in his field and network with other geologists.
Challenges: Alex is still relatively new to the field and feels like he has a lot to learn. He often feels overwhelmed with the amount of data he has to analyze and has difficulty making decisions without more experience. He also struggles with communication and collaboration with his team and often feels like he lacks the necessary soft skills to work effectively with others.
Name: Dr. Catherine Lee
Age: 47 (20 years of experience)
Occupation: Senior Geologist
Education: PhD in Geology
Goals: Dr. Lee's goal is to continue to work on challenging and innovative projects that contribute to sustainable development. She wants to inspire the next generation of geologists and help develop their technical and leadership skills.
Challenges: Dr. Lee faces challenges in keeping up with new technologies, managing teams with diverse personalities and work styles, and balancing technical and managerial responsibilities. To overcome these challenges, she needs to continually learn and adapt to new tools, ensure her team is working collaboratively towards project goals, and make informed decisions that balance technical requirements with effective leadership. These challenges are essential for her to succeed in her role and contribute to sustainable development.
Solution & Impact
Leveraged Watson Visual Recognition to develop an AI-based product that classifies 2D rock images based on their visual attributes, providing petrographers' expertise in preservation, and enhancing precision, speed, and consistency while producing deterministic data that leads to more reliable subsurface models. The successful deployment of the product allows ADNOC to evaluate up to 25,000 images each day, with a processing speed of 527 per second, boosting productivity by freeing up skilled professionals for high-value tasks. Notably, relying on senior or intermediate petrographers would restrict the classification of a maximum of 50 or 25 images a day, making this a significant achievement.
Contribution
I have extensive experience in building computer vision models by collecting images, identifying important attributes, splitting data into training, testing, and validation sets, and developing ML models on IBM Watson.
My ability to work closely with stakeholders played a crucial role in the success of product development and launch. Specifically, I regularly updated the stakeholder on the model's performance and changes and conducted knowledge transfer workshops to help the client operate and retrain the models. Additionally, I collaborated closely with the client's senior geology specialist to incorporate their subject matter expertise into the visual recognition model. By doing so, we were able to ensure a high degree of accuracy in the model's classifications of different types of lithologies. Through ongoing discussions and feedback sessions, we were able to fine-tune the model and ensure that it accurately reflected the geology specialist's knowledge and expertise. This level of collaboration helped to build trust and credibility with the client and ensured that the final product met their specific needs and requirements.
I was responsible for creating the UI design and validating wireframes with the client. I worked closely with the front-end developer to deliver our vision and conducted manual QA to ensure the product is functioning as expected.
Finally, I presented the product at the 2018 ADIPEC conference to discuss the application of AI in the Oil and Gas industry. With my expertise and dedication, I am confident that I can assist you in developing a successful visual recognition model for your business.
