Digital AI Disruption
With a passion to solve the challenges facing the energy industry, experts from both the energy and artificial intelligence sectors joined forces to create OriGenAI. Our mission is to push the boundaries of AI in the energy sector by producing and commercializing disruptive research. Our platform predicts and optimizes complex processes crucial to energy supply with world-leading performance. More efficient energy production is not only a matter of money but also critical to our shared environment.
Many of the most challenging optimization problems cannot be solved with conventional approaches. The problems facing the energy industry are more complex and dynamic than ever before. Our team has developed unique, deep reinforcement learning algorithms which have been used for real-time optimization and planning across multiple domains, and have yielded unprecedented results.
Our new deep learning architectures substantially improve prediction within complex chemical industrial processes. We can predict the outcome of complex industrial processes several hours in advance, allowing for better-informed predictive analytics to drastically increase the efficiency of industrial processes. Better decision-making improves performance and saves millions of dollars in energy costs.
We pioneered the use of Generative Adversarial Networks (GANs) to augment datasets. By learning from essentially boundless datasets, we implement optimal solutions for unforeseen situations. We lead the market in analytical techniques to better predict the precise outputs of refineries, wind farms, etc. allowing energy companies to lower the cost per barrel or megawatt.
1 Provide robust predictive analytics tools to increase efficiency
2 Optimize refinery controls system to increase efficiency
3 Reduce operational cost
1 Optimize field development plan in real time (adapt locations and well control optimization based on on-line information)
2 Provide robust reservoir predictive models that don’t require expensive reservoir simulator tools
3 Use predictive analytics tools to control and monitor oil reservoir wells
1 Optimize turbine controls (including yaw, pitch, angle, and brake) to maximize output at system and farm levels
Read Case Study
One of the main challenges of the downstream industry is to provide a predictive analytics tool that forecasts the behavior of one machine with enough time to reduce operational problems and increase efficiency.
OriGeniAI predictive analytics platform has been tested at the Petronor oil refinery located in Spain. Thanks to our forecasting tool, Petronor is saving millions of dollars at each refinery.
OriGenAI is pushing the boundaries of AI, developing disruptive research and platforms that solve the new challenges of the energy industry.
CEO & AI expert in Oil & Gas
AI energy expert with 10 years of experience leading projects as principal investigator in AI and Oil and Gas at IBM and Repsol. Ruben has more than 15 patents, tens of publications and was recognized by the World Economic Forum as a leading subject matter expert in AI and Oil and Gas in 2017. He was the principal Artificial Intelligence advisor for Repsol and IBM during the last 4 years, leading the digital transformation and several research projects such as (Excalibur and Pegasus).
AI expert in videogames
Julian Togelius is an Associate Professor in the Department of Computer Science and Engineering, New York University. He works on various aspects of artificial intelligence, such as stochastic tree search, hyper-heuristics, deep learning, evolutionary computation and reinforcement learning. These methods are applied to and tested on various domains, with a specific focus on games, which pose a number of interesting problems for AI. His current main research directions involve search-based procedural content generation, general video game playing, player modeling, and fair and relevant benchmarking of general AI. He is the Editor-in-Chief of the IEEE Transactions on Games. Togelius holds a BA from Lund University (Sweden), an MSc from the University of Sussex (UK), and a PhD from the University of Essex (UK). He has previously worked at IDSIA in Lugano (with Jürgen Schmidhuber) and at the IT University of Copenhagen.