- This month, we spent time with Jason Roe, CEO of Avitas, a Baker Hughes venture.
- In this interview, Jason sheds a little light on Avitas, how it operates as the first venture for Baker Hughes and as a technology accelerator.
Q: What is Avitas in a nutshell?
We are a Smart Inspection and Monitoring Solutions company with our primary focus in 3 areas, Automation Solutions, Environmental Solutions and AI Solutions.
The easiest way to think about what we do is: we help to automate manual processes in our customers’ operations. Through the use of machine learning, computer vision and/or robotics, we can take a very manual process and automate it.
For example, at present, to complete a well pad inspection in the Permian Basin, someone today would have to drive around well by well, and literally listen for leaks, look for visual defects – and cover as many sites as possible in a day. With robotics and powerful analytics, we can instead use a drone that will flyover many well sites in a single day, ingest all the data and produce information using computer vision, processed with our own algorithms to provide automated insights to operators into what’s going on with their asset in a much quicker and more efficient way.
Q: What are the benefits for the operator of automating this inspection process?
I think there are three clear immediate benefits for using Avitas’ for this activity:
- One, it reduces costs and adds efficiency: it can materially reduce the people needed to carry out the physical inspection schedule, and it can turn these people to more added value activities like fixing problems instead of searching for them.
- Second, it’s a safer way to operate, it means less ‘windshield-time’. That’s people behind the wheel inspecting sites who can instead be, utilizing the automated findings, and preparing maintenance planning to ensure production uptime. This also takes people out of harm’s way for specific inspection activity on the wellsite.
- And last- but not least, environmental impact: with less driving, the reduced consumption of fuel is immediate versus battery powered drones with higher efficiency.
This is just one area where we operate.
Q: Beyond drones, what other technology are you able to deploy?
Two other areas where we play is in Environmental and AI Solutions.
Inside of Environmental Solutions is the Lumen Product Line with both the Lumen Terrain and Lumen Sky offerings. These offerings help our customers reduce their emissions through mobile and fixed/continuous environmental monitoring and inspection solutions. We currently are focused on methane detection, localization and quantification but in the future could include other VOCs.
Inside of AI Solutions we capitalize on our core competences in Machine Learning (ML) and Computer Vision. We utilize these specialties to offer a service-based solution to either supplement or complement an existing product in the Baker Hughes portfolio or as a service to a customer offering predictive insights with our ML offering, or automated visual inspection enhancement through our computer vision offering. An example is how we are currently using the Computer Vision expertise to supplement the flare.IQ product offering for a close loop solution.
Using the core skill sets we have in data science can bring material value to both our customers and as a supplement to some of the existing Baker Hughes portfolio offerings.
Q: Since this is the first of Baker Hughes’ ventures, how do you work with the rest of Baker Hughes’ business?
Let me give you a couple of examples of how it’s very helpful to be part of the Baker Hughes portfolio.
First, we are able to leverage the broad infrastructure- the many facilities around the world, and the expert sales team. For example, we work from a local site in the Permian Basin to deliver to our customers in the field. That way, we can be in touch with the right stakeholders locally, close to our customers.
Second, as mentioned above, we provide our services as additional insights, augmenting Baker Hughes existing services.
Both the Lumen and flare.IQ examples highlight how we can contribute to existing products and make them better for our customers. In addition, we continue to work with some of the other Digital Solution offerings to see if we can add ADR (automated defect recognition) to existing inspection solutions through our computer vision platform to automate some of the manual work from the inspections alone.
Our aim is to provide the additional service only where it would matter to the customer to supplement a product they are already using from the Baker Hughes suite.
Q: What does it mean to be a venture company?
Being a venture, primarily enables us to operate a little more autonomously than the rest of the Baker Hughes business. We are fortunate in that we have access to the scale and expertise that the Baker Hughes portfolio and infrastructure possesses, while also being able to maintain some of the entrepreneurial spirit and culture that an upcoming company must possess.
Being able to pull in resources at times enables us to gain quicker credibility and access to customers and support, but we also think a little like running as a start up and not taking it through all the processes of the larger organisation within the group.
By being agile and trying to keep the entrepreneurial side of the business thriving: this filters down to how we manage our customer relationships, how we engage the supply chain, and fitting in within the Baker Hughes structure.
Q: Can you tell us about a specific project you are working on?
Currently, we operate mostly in the US for aerial inspection, and starting in other regions such as Saudi Arabia for other projects.
With a customer in the Permian Basin- we do well-pad inspection.
The pilot programme includes 30 to 40 well pads to fly, run analytics, and produce a report by 7am the next day, when their staff have their morning review to plan the day ahead.
We flew all the sites, multiple times, twice a week and provided automated reports in advance of their 7am staff meeting. These reports highlighted issues on each pad: a leak here, debris on site here, etc.
For another operator in Prudoe Bay – we are taking 30 years of operational data of their pipeline network there. construction data, inspection data, environmental data…etc.
The customer sent us huge databases of information and we ingest it, fuse it, inject our machine learning algorithms and physics-based models. We can then give them predictions on wall loss, corrosion under insulation on their pipeline network, amongst other things.
We are able to highlight high and lower risk areas, percentage of certainty of an issue occurring at a particular time: the customer can in turn refine their inspection programmes, making sure they hit the areas of concerns based on the data analysis- that might be of real consequence.
For more information on Avitas technology and projects,