San Francisco startup Pratexo is helping companies overcome the traditional technical hurdles of complex, edge-to-cloud deployments. The startup’s no-code platform significantly accelerates the ability of engineers to design, test, deploy, and manage complex, decentralized software architectures from the far edge up to central clouds. The result is reduced time to deployment, improved security, reduced energy consumption due to improved system efficiency, and the ability to run applications and analytics while disconnected from central servers.
Blaine Mathieu CEO Pratexo, San Francisco
A former Gartner analyst and founder, CEO/CMO/CPO at public tech giants and startups, Blaine Mathieu is based in the San Francisco Bay area. He is CEO of Pratexo, an edge-to-cloud solution acceleration platform company.
For more information, visit www.pratexo.com
ABB Review (AR): Pratexo’s focus is edge-to-cloud computing – an area that is all about decentralization. To put things in perspective, what’s driving the decentralization trend?
Blaine Mathieu (BM): Specific to the electrification space, meaning areas such as grid operations, EV charging networks, and virtual power plants, the drivers include resiliency, flexibility, and a reduction in overall costs via increased system efficiency, along with the ability to potentially sell unused power capacity back to power markets. But the general trend toward decentralization is relevant to virtually all industrial systems. It’s about moving away from monolithic structures and systems towards real-time operations that can only be optimized at the most local level. The last 20 years of IT have been mostly about centralizing computing into the cloud. The next 10 will be about balancing that with a hybrid edge-to-cloud approach – doing the right compute at the right place and at the right level. Accelerating that transition is what Pratexo fundamentally enables.
AR How does this relate to sustainability initiatives?
BM There is a direct relationship →01, and we are definitely seeing more market pull from this area. More efficient, available, and safe systems are the foundation of most sustainability programs. That precisely mirrors the value that decentralized systems, running in real time and adjusting to rapidly changing circumstances, can provide. Since we focus particularly on improving the operations of industrial and infrastructure systems, we have positioned Pratexo as a Catalyst for Sustainable Change.
AR Companies are facing the double challenge of balancing a steadily increasing level of data creation and rising system complexity, while driving their transformation from basic analytics towards real-time, highly efficient, autonomous system control. How does the Pratexo platform support customers in this evolution?
BM Pratexo’s goal is to help companies deploy software solutions faster and with lower project risk by giving them the tools to overcome the traditional technical hurdles of complex, edge-to-cloud deployments. The Pratexo Studio is what enables this: our no-code platform significantly accelerates the ability of an architect or engineer to design, test, deploy, and manage complex, decentralized software architectures from the far edge up to central clouds. The Studio builds, deploys, and configures complex multi-tiered architectures via a drag-and-drop interface to quickly create hyper-connected networks that enable data collection, advanced analytical processing, and real-time applications at scale. The result is reduced time to deployment, operational continuity in the event of central server disconnection, and a more secure, reliable way to manage distributed systems, including power grids, EV charging networks, distributed manufacturing operations, and many other areas of industry and infrastructure.
AR In practical terms, what do these capabilities mean for software architects and solution developers?
BM It means they can get projects designed, tested, live, and in the field in a fraction of the time it would otherwise take. As a proof point: Pratexo now has a global framework agreement in place with the Cognizant IoT group. The reason is, as one of their most senior leaders told us a few months ago: “What you demonstrated in 10 minutes would have taken a team three months to do.”
This is also the fundamental reason why Pratexo won the ABB Electrification Startup Challenge back in May 2022. While most entrants were able to create a compelling vision and slide deck over the 10-day period of the Challenge, Pratexo was able to use its platform to stand up an actual smart grid simulator running advanced analytics in real time, on top of a micro cloud connected to real sensors and devices. That’s what is possible with a platform like this.
AR How does Pratexo Studio work?
BM In the first step, the architect begins to define the sources of data and associated software components that are necessary to support the analytics and applications that end-users will benefit from →02. They can decide where each component needs to be deployed, perhaps on each compute node running at the system’s edge, or maybe onto a Kubernetes-based micro cloud that runs across a set of compute nodes. This has the advantage of being able to run advanced applications and analytics locally, even if the system is disconnected from central clouds or servers – which is often a requirement for mission-critical systems.
Of course, these solutions can also be deployed all the way up to central hyperscaler clouds – after all, this is what edge-to-cloud computing is all about! In fact, before any software is deployed to physical environments, Pratexo enables the deployment of a digital twin of the system to a virtual environment in the cloud to begin simulation and testing before a single hardware device or compute node is provisioned →03. Pratexo is device, sensor, network, compute resource, and cloud agnostic. We don’t provide infrastructure – we enable the rapid deployment and management of solutions on top of infrastructures.
AR How does a customer get from an architecture design to a finished solution?
BM Some companies can take an edge-to-cloud architecture that the Studio deploys and very quickly begin building solutions on top of it. But many benefit from additional support. Pratexo does this in two ways. First, we have developed what we call ’Solution Frameworks’ that provide a kick-start towards implementing highly customized solutions on top of the Pratexo platform. Currently, we have solution frameworks for integrated system monitoring and alerting, and one for root cause analysis expert systems. We are now in the process of developing further solution frameworks.
Second, we have a professional services group that collaborates closely with our clients and partners to either take a Solution Framework and complete the necessary customizations, or to develop a complete solution from scratch that is accelerated by the platform. Whether it starts with a Framework or not, this approach is powerful because it is almost impossible for out-of-the-box solutions to meet the very specific needs of each customer and use case. Over time, we definitely see third parties such as systems integrators and engineering, procurement and construction companies (EPCs) also being able to use our platform and Solution Frameworks for their customers – which explains our partnership with Cognizant that I mentioned earlier. Finally, we are excited about close partnerships with OEMs and solution providers like ABB that are working on wrapping more complete solutions around their equipment offerings, some of which may be delivered as-a-service. This is the basis of much of our current work and ongoing discussions with teams at ABB.
AR Although Pratexo is still a startup, does the company already have customer success stories?
BM Yes! A great example is Hallingdall Kraftnett (HKN), an innovative Norwegian power grid operator that is seeking new ways to improve and safeguard its efficiency and system uptime →04. Like all operators, HKN is faced with power spikes and rapid changes in demand – particularly based on the large loads placed on the Norwegian grid for recharging electric vehicles, which make up the majority of cars on their roads today. While this influx of EVs is helping Norway lead in some measures of sustainability, grid operations are hampered by dark data, which is data that is not actually used to drive operations. Operations are also hampered by inadequate communications between remote substations and prohibitive cloud computing costs. The company operates thousands of substations that collectively generate many gigabytes of data every hour – far more than could be pushed to a central cloud.
HKN and its partner grid operators are deploying edge nodes at each substation →05. They have also grouped certain nodes into regional micro clouds that are able to share compute resources, bringing the power of cloud computing down to the grid edge. For the first time, the massive amounts of data the machines in each substation are generating – including legacy PLC data and data from ABB sensors and equipment – are being analyzed, and events processed, in real time.
Now that the platform is in place, new solutions and analytics are easily provisioned. This includes microphones with an associated algorithm that can listen for the sound of a partial discharge (PD) at each substation in order to alert maintenance teams more quickly than otherwise possible using typical manual inspection protocols. Future use cases are focused on improving the physical security of remote operations. No longer are all these systems and solutions stuck in silos.
AR Pratexo is working with ABB to create a type of expert system. How will this support decision making?
BM This is an exciting project for Pratexo. One of the Solution Frameworks we used at the ABB Electrification Challenge was our Root Cause Analysis Expert System Framework →06-07. The purpose of this Framework is to accelerate the decision-making capability of human experts in their quest to understand why complex systems of machines fail or do not run optimally. Since the Challenge, we have been working with experts at ABB to create Expert System Solution Modules that run on the Framework. These Modules contain the custom data ingestion extensions, rule sets, algorithms, and reporting templates used by the Expert System Framework and are specific to each type of machine/asset or set of machines.
These expert systems can be run either in forensic – meaning retrospective – mode, or potentially in real time, streaming mode and can be run anywhere from a micro cloud on the far edge, near the machines in question, up to central clouds. The results include detailed analytics of what could potentially be causing machine faults. Although this project is still in early development with ABB, we are very confident about its ultimate value in the market.
AR Any final thoughts on Pratexo’s collaboration with ABB?
BM Since ABB’s SynerLeap innovation growth hub introduced us to ABB in early 2022, it has been a true pleasure to work with the individuals and teams there. In addition to the specific projects I’ve mentioned, we now have half a dozen discussions ongoing with other groups at ABB, and we truly look forward to accelerating more innovative and sustainability-oriented projects with our partner in the years to come.
Photo right: ©C.Castilla/stock.adobe.com
