The search for oil and gas involves huge amounts of data. Interpreting these datasets is repetitive, prone to human error and bias, labour-intensive, time-intensive, and expensive. Furthermore, the amount of data is too vast for a human to make sense of using conventional methods, which means that a large proportion is not analysed and therefore its potential value may be overlooked.
We wanted to find a way to extract value from these complex data. To do this we use modern machine learning methods to allow new insights into big oil and gas datasets. Through the building of proprietary workflows and software we can:
Process high-dimensional datasets to allow valuable insights into previously overlooked or impossible to process information
Increase the efficiency of geologists or engineers by about 40% by automating repetitive tasks
Significantly reduce human-bias and error from data analysis
The idea behind Latent Analytica came from a research collaboration between earth scientists and Engineers in the faculty of Earth Science Engineering at Imperial College London. The team are all studying in the faculty and we met during research seminars. We are currently at the stage of product development.
Advisors and mentors
Dr Graham Ganssle, Head of Data Science at Expero, is our advisor who became involved through industry collaboration.
Enterprise Lab Support
We have taken part on the Venture Catalyst Challenge (VCC) and are now part of the Imperial Venture Mentoring Service. Both of these programmes have kept our team focused on building and developing the best product possible.
Successes and setbacks
So far our biggest successes have been getting through to the semi-final of VCC, building a team with a diverse range of skills and presenting our work to leading companies in the industry.
In terms of setbacks it was a blow when we didn’t reach the finals of the VCC and we have also missed out on funding opportunities. Also the slow progress of our development has been difficult at times when we would like to move forward faster.
Advice to aspiring entrepreneurs
Build a diverse team with a broad range of skillsets