AWS, IBM and ElectrifAi announcements at Adipec
- November 15, 2021
- Steve Rogerson
This week’s Adipec event in Abu Dhabi saw major smart energy announcements from New Jersey-based ElectrifAi, IBM and Amazon Wed Services (AWS).
Hosted by the Abu Dhabi National Oil Company (Adnoc), the Abu Dhabi International Petroleum Exhibition & Conference (Adipec) is a meeting place for oil, gas and energy companies and professionals.
ElectrifAi, a specialist in practical artificial intelligence (AI) and pre-built machine learning (ML) models, announced computer vision (CV) and machine learning-as-a-service (MLaaS) for the oil, gas and energy industries.
IBM and AWS are combining IBM Open Data for Industries for IBM Cloud Pak for Data and the AWS Cloud to serve energy customers. This is built on Red Hat OpenShift and will run on the AWS Cloud, simplifying the ability for users to run workloads in the AWS cloud and on-premises.
Oil and gas companies need to leverage the power of AI, ML and CV to drive operational and cost efficiencies. They generate large amounts of data, but frequently the value and power of those data are never fully realised due to the inability to access siloed data, general lack of data engineering and data science talent, and an inability to connect ML and AI to the practical needs of business units.
ElectrifAi has one of the largest libraries of pre-structured ML models that have been built and battle tested over the past 15 years. The company has also developed CV models to drive workplace safety as well as reduce costs.
And now with the MLaaS offering, ElectrifAi enables companies to realise the benefits of AI and ML. Clients simply describe their business use case and ElectrifAi tells them what data are required. Clients then supply the data and ElectrifAi takes over from there by training, operating and deploying the models and delivering results quickly.
In the oil and gas industry, human error in high-risk environments can cause accidents, loss of life and production stoppages. CV can play a role in helping prevent injuries, keeping rigs and production facilities running smoothly by alerting operators to potentially dangerous situations and remediating issues before an accident occurs. Similarly, CV can also monitor valve banks and other critical infrastructure for human error as well as wear and tear leading to equipment failure. In addition to losses from production stoppages, there are also ancillary costs that can be avoided such as litigation, regulatory fines and increases in liability insurance coverage.
MLaaS increases the efficiency and convenience of machine learning. Clients can get started with machine learning without installation processes or providing their own servers. Although most companies are actively exploring the possibilities of AI and ML, many struggle with basic challenges around data availability and quality as well as being able to recruit and retain data engineers and data scientists.
ElectrifAi’s MLaaS offering addresses these challenges. With MLaaS, companies need little to no experience to realise the business and operational benefits of AI and ML. MLaaS deploys easily within any cloud environment or on the user premises. ElectrifAi will also develop, operate and maintain the models on behalf of the client.
By using MLaaS, companies can achieve benefits to improve their operations and capabilities, including lower costs, as one model can be cheaper than the annual cost of a single data scientist; faster time-to-deployment and lower project risk, as the average deployment is between eight and twelve weeks for MLaaS versus eight to twelve months to build new ML models; and faster time-to-value with a high return on investment.
One of ElectrifAi’s MLaaS offerings is ML-based spend and procurement analytics that have been trusted by some of the world’s largest companies. It is quick and easily deployed in a client data centre or any cloud environment. Average realised savings for the SpendAi product fall in the 2-4% range.
“We’re pleased to introduce our computer vision and machine learning-as-a-service offerings to the global energy industry,” said Edward Scott, CEO of ElectrifAi. “Every company can now easily achieve the benefits of computer vision and machine learning with a very high RoI. We are helping energy companies across the globe grow and become more competitive through data-driven business decisions.”
The energy industry is also facing pressure to reduce greenhouse gases as demand for affordable energy continues to rise. Energy companies need to drive efficiencies to free up capital, time and resources to invest in discovering more sustainable energy sources. Data and digital technologies can help to navigate this transition, yet an IBM survey found that less than half of oil and gas executive respondents are using data to drive that innovation. This is in part because most digitisation efforts have been in proprietary closed systems, hindering the potential to combine and increase the value of data.
The collaboration between IBM and AWS aims to accelerate reduction of data barriers in the industry. IBM Open Data for Industries is open source using the OSDU data foundation for the oil, gas and energy industry. It is integrated with IBM Cloud Pak for Data for easy data management, and built on the Red Hat OpenShift Kubernetes platform and open architecture, designed so companies can run and operate applications universally.
With this collaboration, users will gain the flexibility to run OSDU data platform applications in the AWS cloud or on-premises while addressing data residency requirements. Combined with the cloud infrastructure of AWS cloud services, this data platform can help energy companies reduce the cost, time and resources needed to leverage data to derive insight, streamlining operations and transition to sustainable energy generation.
“Much of the data needed to solve the complex energy challenges, such as superior subsurface decisions, already exists, yet is untapped,” said Bill Vass, vice president of engineering at AWS. “This is because one of the greatest values of those data is derived when they can be effectively combined, but usually these data are locked by data residency requirements, legacy applications or proprietary data formats. By collaborating with IBM and leveraging Red Hat OpenShift, we will be able to offer customers a global, seamless offering with the flexibility to run on virtually any IT infrastructure and drive longer-term digital innovation.”
The OSDU Forum is a cross-industry collaboration that provides a vendor-neutral framework for companies to develop data platforms against common energy industry standards. By working together, IBM and AWS aim to accelerate the value of this platform for global users. The goal is that this combined effort will help serve the needs of energy companies today with flexibility to adapt to change amid energy transition.
“Data are a critical asset to help fuel energy transition, yet too often energy companies must choose between running applications on-premises or in the cloud, and often each deployment uses a proprietary data format,” said Manish Chawla, global managing director for energy, resources and manufacturing at IBM. “This means that rather than using all of those collective data to gather insights, augment operations and inform innovation, some of them were going unused. Our collaboration with Amazon Web Services is addressing the need to make it easier for energy customers to access their data and provides the industry with a flexible solution to meet the challenges of today, as well as more easily adapt as the industry evolves.”