1. Offer of Remote Monitoring of Fuel Tanks & Storage


Tanks and storage are one of the oldest, most fundamental industries to mankind, and the most recent to see a revolution through IoT Technology.  At the dawn of agriculture, tanks and storage allowed humans were able to store food during throughout the year, initiate commerce, and protect their assets.  Today, this age old technology helps protect everything from fuels and fertilizers to clean or waste water management. Innovators in the industry are using digital-twin IoT Technology to revolutionize this space 

Protect from theft.

Theft in the industry is rampant.  Victims include owners of personal storage tanks as well as industrial size tanks.  Siphoning assets from the tank is commonplace, with the Associate Press reporting that Mexico lost up to $3.5 billion in 2018 from siphoning of its oil tanks and pipes.  With IoT technology, real-time monitoring, and machine learning algorithms, you can know when someone is tapping into your assets. Algorithms that check abnormalities in flow, pressure changes, and other factors can indicate that someone is redirecting your investment elsewhere. 

Monitor the Linings. 

The steel shell of a tank is not the only critical element of the structure. Tank linings, often made with polyurethane or PVC, can be compromised too, and often need regular changing.  Using IoT technology, you can find out when your lining starts to leak, preventing valuable assets from being lost, as well as maintaining safety and regulatory compliance for your organization.

Watch for overflow.

Overflow is a hazard to be remote-monitored as well.  Tanks may be filled improperly and the dimensions of the tank can change with season and temperature.  An overflowing tank can cause leaks, loss of assets or even lead to fires. Remote-monitoring, standardized data collection and machine learning can help you better understand how third parties are interacting with you tanks, so you don’t need to be there monitoring in person.

Monitor for depreciation in the steel, leaks with EDGE-IoT.

The tanks themselves change slightly with the stress of seasons and weather. By using EDGE-IoT & AIoT (Artificial Intelligence of Things) connectivity, you can monitor the change in volume of the tank to see if the changes are attributed to cyclical changes in the tank because of temperature, or if a more hazardous leak has occurred. 

Watch for degradation in the contents of the tank.

Contents within the tanks can change over time.  Instead of sending a diagnostic technician to take samples year-round with expensive lab tests, use a real-time remote monitoring to check to see whether the what you put in the tank months ago, is what you’re getting back out of the tank when it comes to delivery.  Factors like climate change, seasonal temperature variation, or small contaminants can all affect the value of what is stored in your tanks.


2. Industrial benefits of predictive maintenance

Predictive Maintanace Value Graph_WC.png

Ekatra’s Predictive Maintenance Platform reduces your working capital costs, and helps smooth out your business

Predictive Maintenance is a cornerstone of Ekatra’s platform features, and is rooted in our proprietary machine learning techniques. Below are 4 ways the Ekatra’s Machine learning platform uses real-time sensor data to help you stay in control of your business, and manage working capital expenditures.

  1. Isolating local machine failures in real-time. When you can quickly identify and isolate a part of the machine that is about to break, that prevents the “cascade of consequences” that can lead to a system wide failure. Real-time sensor analytics send data through the EDGE-IoT network a while family of models that can explain what might be going on. Multi-class and unsupervised models help us to not only identify issues that have been encountered before, but also try to isolate issues that may ever have been seen (through Unsupervised machine learning techniques).

  2. Identifying what parts will fail in the future. By using real time data and studying failure events as they come about, you can predict when those parts are likely to fail in the future. Anomaly detection and regression modeling help us to dig deeper into issues we are studying. This reduces the likelihood of disruption to your supply chain and manufacturing lines.

  3. As a trustful quality watchdog, Ekatra ensures your partners are following their Service Level Agreements, and making sure parts fail when they are supposed to. Breakdown is an inevitability of in any engineering endeavor. However, whether it breaks down in a predictable way can mean the difference between what is covered under warranty, and a multi-billion dollar recall.


3. Offer of Compliance & Safety via Electrical Grid Assets


By looking at the value shifts in the remote monitoring and electrical grid industry industry, we identify four ways Ekatra has already helped its customers, by avoiding a costly Cascade of Consequences -- helping your business move into the future. This is the passive way to raise your quadruple bottom line, without worry about receiving a phone call in the wee hours of the night, to put out fires, both metaphorical and actual:


       Your IoT monitoring system is your 21st century insurance policy against operational disruption, which immediately affects your bottom line. 

We saw this value in Ekatra’s EDGE-IoT Technology public utility case study of Consumers Energy. The ability to predict and navigate potential operational issues, like predict if a transformer will stop operating, enables you to be pro-active in the face of impending issues and plan around it, instead of reacting after the damage is done and questions need to be answered. EDGE network technology also monitors the cascade of consequences, and signals the appropriate stakeholder in your company to make sure the right person’s phone rings to resolve the issue as quickly as possible.


       IoT can help delay depreciation of your tangible assets. In Ekatra’s Case Study of Transformer and Critical Asset Depreciation, we identified three immediate factors where IoT sensing helped actually create value to the balance sheet. Most depreciation tables are not dynamically adjusted at all from the day they are put in place. Thus, there is a missed opportunity to adjust those tables with real-time information to capitalize on if the asset is operating better than it otherwise was expected.


       Through Digital Twin technology & case study of critical infrastructure, management teams do no need to worry about receiving a call at 2 a.m. to delegate how a break in the system needs to be managed. With EDGE-IoT technology, the fabric of the IoT technologies are continuously in communication and accountable to each other, efficiently notifying only the relevant stakeholder when issues arise, instead of sounding alarms to the whole company. This means that if a sensor senses something wrong, IoT systems are designed to know who next in the chain of communication needs to be informed, isolating the potential damage and leading automatically to contacting the technical person who is responsible for initiating the immediate work.


In the age of digital social media, repetitional risk is just as much a cost risk as a defective device. When it comes to Twitter for example, the greatest risk to companies and the confidence in their service is for consumer to be ahead of the company when it comes to critical news, like a power outage or a safety hazard nearby. The end consumer is almost always aware of issues in real-time, like when the power goes out in a residential house, or your cellular service breakdown, but it is the delay in minutes between the consumers real-time knowledge and your companies awareness that determines how much repetitional loss you have. In fact the only way to guarantee minimizing that loss is to become as real-time aware through sensing as possible.

       In fact, Consumer Power has turned social media into a positive factor in their business, immediately alerting those who might be affected, and even documenting the hard work of their workers who brave weather events and hazardous conditions to help their customers.

Ultimately, to prevent the Cascade of Consequences, compliance needs to be thought of as a necessary pro-active part of every enterprise built for the future.


4. The Offer of Quality Assurance & Managing CapEx

Predictive Maintanace Value Graph_CapEx.jpg

USing Ekatra’s Machine Learning Platform to tell you what you own and what its worth

In your business, the life of a machine is an investment with a finite amount of life. Unless you are Disney Corp, where you renew your trademark on Micky Mouse every 70 years without fail, machines with moving parts have a life and value that amortize and depreciate with time. At Ekatra, we use Machine learning and AIoT you can reduce the capital expenditure on moving parts in your manufacturing supply chain.

  1. How can I maximize the life of the machine, and the integrity of my value chain? In engineering, we are constantly concerned about the “Unknown-Unknowns”, or those things that can break, but we have no idea “if”, “how”, or “when”. The Ekatra Platform has a ecosystem of machine learning algorithms that helps you discover never-before seen outcomes, because they disrupt your business and cost you Capital Expenditure. For example, if the timing rod in an engine with not correctly designed that could yield issues never before observed. When trying to isolate a problem quickly, separating new issues from those you have observed before is critical. Like a heart with a murmur, it is often hard to tell where the murmur is coming from and what it means. Unsupervised learning algorithms combined with real-time data, help to paint pictures of issues that may have never been observed before. This means that the way that a device fails, can continue to be predictable and safe.

  2. Managing Capital Expenditures by my managing your “Known-Unknowns”. Known-Unknowns are things you know that are going to inevitably happen, but aren’t exactly sure how or when. For example, it is inevitable that a car engine will eventually breakdown, however, it is more challenging to decide at what millage. Machine learning and measuring a device in real-time helps us to better understand not only when that will happen, but also what the value of the device will be once that issue occurs. When making executive decisions, machine learning and regression analysis can be key in maximizing the value of the machine toward the end of its life.


5. The Offer of Optimization

Optimizing is about Attacking Bottlenecks.jpg

Ekatra’s Optimization is not a result, it is a process that enhances your whole value chain

Optimization is only has good as your tightest bottleneck. At Ekatra, we have identified 6 major stages in your vertical stack where we help you create value by reducing inefficiencies in the manufacturing and feedback process.

  1. Sensor Level Connectivity. By using EDGE-technology, we help ensure that redundancies and continuity is in place at the very source of your data feed, at the sensor level. Our data-feeds collecting data points at over 500,000 per second, help to make sure that you are in tune with every breath of your system.

  2. Data Management Layer. Sensors can deliver 500,000 data points per second, however, without a system to collect and manage that data, the value delivered by the sensors is lost along with the data resolution. High speed, real-time data collection, like with Spark, or using meta-data layers from digital twins, can help to manage data at all levels, so it is stored and managed with integrity.

  3. Digital Dashboard with Human Machine Interface. At Ekatra, we know that you can’t look at 500,000 things at once, and you need the help of a Dashboard to see what is important, when it is important. Our experience with Nuclear Reactor Sites is a prime example, of the need to create a dashboard that doesn’t tell you everything right now, just the life-death meta-factors that you need to know right now.

  4. Using machine learning to help guide your view of the system. Machine learning is good at checking whether a system is behaving as we expect, which means it’s also great at guiding our attention toward what we didn’t expect to see. Machine learning can reduce the time required to identify issues, as well as help train your digital twin to evolve as you collect more data.

  5. Third-party business applications are the foundation of any flexible platform. While Ekatra’s valuable technology is proprietary, our API is for public, good-will use to anyone who wants to develop a third-party app. By opening our platform to third party developers, the number of solutions available to our clients opens approaches the infinite.

  6. Modular system components are the foundation of Ekatra’s ease of use & maintenance. Modularity means that your system can be protected and scaled. Whether you are looking at a small local system or a system that spans time-zones, modularity means that not only is installation consistent and predictable across a system, but security after installation is uniformly empowered across the board.


6. The Agricultural Industry & Multi-factor flavor enhancing


EDGE-IoT is optimizing agriculture with hierarchical models to enable us to optimize, while in the middle of a growth cycle

Agriculture presents its own challenges when it comes to creating the best product at an industrial scale. When manufacturing an inorganic product in an industrial setting, like a car engine, the process is largely deterministic. This means that the inputs are clearly identified, like steel as raw material and metal cutting machines to forge the engine block, and you may have an objective idea about how the engine is going to perform in the world.

However in agriculture, the flavor of fruit for a person or the effects of cannabis when consumed are largely subjective, requiring a human to evaluate. A computer cannot yet model the human tongue. Additionally, not only is the “performance” of agriculture a more random process, but the inputs, like light from the sun, are stochastic as well.

There are three important reasons why the use of Bayesian models are critical in agriculture and the future of IoT.

  1. Bayesian models are good at handling inputs that change. Since there are a lot of conditional inputs used in agriculture, like sunlight, how light hits the leaf, water absorption or soil nutrients, you need models to model your inputs, before you start thinking about outputs. The way inputs enter a plant is actually a stochastic process. Just because you poor 1 cup of water in the soil next to a plant, does not mean that the entire 1 cup will be full absorbed: some water can evaporate before being absorbed, or some water may leak from the container. The way a plant absorbs inputs is just as challenging a model as the way the plant ultimately grows, and having multi-factor models helps us control for all of these input factors.

  2. Hierarchical models are good with subjective outputs. When looking at the performance of an engine, statistical models are good at identifying factors that can be clearly measure by a machine, like engine power and exhaust pollutant composition. However, when your output is subjective, or conditional on the taste of the observer, then Bayesian models help you look deeper at what factors are actually at play, in what will make a strawberry sweeter to a human or a cannabis strain a better experience for a recreational user.

  3. Bayesian models help us correct during the manufacturing process. Unlike with the development of a car engine, plants can be tested for nutrients and performance factors while they are growing, while a car engine can only be evaluated once full assembled. By checking samples of the plant while it is growing, we have the opportunity to discover if the plant has been deficient in a particular time of nutrient or input, and then continue to change the process in real-time using the data we get from EDGE-IoT sensors.


7. The offer of fleet management

Optimization image.png

Machine Learning & EDGE-IoT means that the engine on the assembly line and the engine in your car, are all being cared for in real-time

Before IoT and Machine Learning, a design team and quality team would be two different sides of the manufacturing cycle, when creating your fleet of trucks or cars. Engines would be designed by engineers in a lab to handle all sorts of different “theoretical conditions", like weather temperatures of -100 to +500 degrees Fahrenheit, and race-car performance tests. Then that engine could be placed into a Minivan and assumed it would also be able to handle the day to day of stop and go traffic.

When managing your fleet, the truth is that, every engine and vehicle, like every fingerprint, has tiny differences and imperfections. Although engine pistons are all cut to within microns of each other in precision, differences still exist. These differences can result in issues that are traceable in your fleet, so we can see how different parts might relate to different issues in performance before they end up resulting in lost business, a full-recall or even lost life.

With EDGE-IoT technology and real-time feedback of data from the field to the assembly line, you can see how nuances on the assembly line result in difference out on the field, so manufacuters and Fleet owners in the Industry 4.0 era can react in real-time to any issues affecting their business.

For instance:

  • Sensor vibrations at the assembly line level, might correspond to issues within a whole group of engines. Since we are able to track the engine parts from the cradle to placement into the car, we can see at the assembly line level, what is going on if a group of engines all start experiencing issues.

  • Sensor vibrations at the machine level, can help us to identify potentially defective machines on the assembly line, and take them off of the line before they cause issues for future products.

  • Sensors on-board cars are able to feed data back to the assembly line, so the manufacturer can tell in real time how those fresh engines are performing out on the field. Separate on-board sensors can see how that engine also interacts with different types of cars, giving another layer of data and results that can help engine and car work better together.