MICAN are experts in the Internet of Things (IoT) as well as IIoT (Industrial Internet of Things). We know this technology deeply from end to end. From the factory floor and robotics and SCADA through the networks to the cloud and to the backend applications like artificial intelligence, analytics, and asset management. Regardless of the IoT applications, whether it is smart cities, mining, oil & gas, assembly lines, sewers and waste water, garbage collection, trucking, trains, or airplanes, we have firsthand experience to share with you on your projects.
Here are a few IoT scenarios to consider and to demonstrate how MICAN can serve your needs. The smart pipeline is offered as an example of what is possible for a modern IoT solution.
IoT Connected Pipelines: Networks
Connecting to gas & oil pipelines can be challenging, if not even seemly impossible at times. Pipelines are often buried and located in some pretty remote locations. At other times they do run through or nearby populated areas.
However, this is a older perspective. Newer internetworking solutions provide richer connectivity options. So, connecting to pipelines is much easier now, regardless if it is in urban, suburban, rural, or remote locations. A myriad of connectivity options are now available, all the while being cost effective to make situational awareness and the granularity of sensor data work better and faster.
Network options include the following:Private IoT networks can quality coverage for many kilometres if implemented correctly. Some of them offer huge data rates over shorter distances, perhaps 1 kilometre or less. However, other technologies offer significant distances ranging from 1, 2, 5, 10, 15, or even 20 kilometres if line of sight is available and suitable height above terrain is possible.
Temperature – A pipeline is subject to both internal and external temperatures. The sun might heat up a pipeline and friction caused by the flows may create heat internally. Sensors can measure internal and external temperatures to 1/100th of a degree to reveal deltas and provide evidence of anomalies.
Humidity – Humidity is defined as the amount of water vapour in an atmosphere of air or other gases. The most commonly used terms are “Relative Humidity (RH).
Solar Radiation – Often used in conjunction with temperature sensors, solar radiation sensors detect the sun’s light and measure the nanometer wavelengths of the colour temperature, hours of radiation, angle of radiation, and broadband variations. A pyranometer is a type of actinometer used for measuring solar irradiance on a planar surface and it is designed to measure the solar radiation flux density (W/m2) from the hemisphere above within a wavelength range 0.3 μm to 3 μm.
Pressure – Pressure sensor is a device that senses pressure and converts it into an electric signal. Here, the amount depends upon the level of pressure applied. These sensors make it possible to create IoT systems that monitor systems and devices that are pressure propelled such as pipelines. With any deviation from standard pressure range, the device notifies the system administrator about any problems that should be fixed.
Vibration – Vibration sensors can be used for predictive maintenance of rotating machines such as motors, pumps, compressors and fans.
Chemical – Chemical sensors are applied in a number of different industries. Their goal is to indicate changes in liquid or to find out air chemical changes. They play an important role in bigger cities, where it is necessary to track changes and protect the population. Main use cases of chemical sensors can be found in industrial environmental monitoring and process control, intentionally or accidentally released harmful chemical detection, explosive and radioactive detection, recycling processes on Space Station, pharma industries and laboratory etc. Following are most common kind of chemical sensors in use:
- Chemical field-effect transistor
- Electrochemical gas sensor
- Fluorescent chloride sensors
- Hydrogen sulfide sensor
- Nondispersive infrared sensor
- pH glass electrode
- Potentiometric sensor
- Zinc oxide nanorod sensor
Gas Detection – Gas sensors are similar to the chemical ones, but are specifically used to monitor changes of the air quality and detect the presence of various gases. They are used for the detection of toxic or combustible gas and hazardous gas monitoring. Following are some common Gas sensors:
- Carbon dioxide sensor
- Carbon monoxide detector
- Catalytic bead sensor
- Hydrogen sensor
- Air pollution sensor
- Nitrogen oxide sensor
- Oxygen sensor
- Ozone monitor
- Electrochemical gas sensor
- Gas detector
Of all the gases emitted from a pipeline, methane is considered to be the worse since it cannot be seen nor tasted, yet it is considered to be the second biggest worldwide cause for global warming and the destruction of the atmosphere. So, detecting leakage from pipelines is not only economically smart, but also critical for environmental protection.
Analog to Digital Adapters – Since most pipeline already exist, and they are old builds, they still use analog gauges. New IoT sensors are now bolted-on to these gauges to read values and convert them to digital signals for seamless integration of the old with the new technologies.
Impurity and Contaminate Detection – Sensors can be used to detect foreign objects and material build-ups within the pipeline. Wax build-up can affect flow rates and create undue pressure in the pipe. Detection can provide insight for cleaning the pipeline with a PIG and then measure the cleanliness after the cleaning.
Valve Actuation – SCADA has been monitoring and controlling valves for decades, now IoT integration with digital values for precision accuracy can be performed.
Video as a Sensor – Image sensors are instruments which are used to convert optical images into electronic signals for displaying or storing files electronically. The major use of image sensor is found in digital camera & modules, medical imaging and night vision equipment,thermal imaging devices, radar, sonar, media house, Biometric & IRIS devices.
Infrared – An infrared sensor is a sensor which is used to sense certain characteristics of its surroundings by either emitting or detecting infrared radiation. It is also capable of measuring the heat being emitted by the objects.
Accelerometer – Accelerometer is a transducer that is used to measure the physical or measurable acceleration experienced by an object due to inertial forces and converts the mechanical motion into an electrical output. It is defined as rate of change of velocity with respect to time.
Gyroscope – A sensor or device which is used to measure the angular rate or angular velocity is known as Gyro sensors, Angular velocity is simply defined as a measurement of speed of rotation around an axis. It is a device used primarily for navigation and measurement of angular and rotational velocity in 3-axis directions. The most important application is monitoring the orientation of an object.
LiDAR – Drones flying over the pipeline, equipped with LiDAR (Light Detection And Ranging) inspection is a better, faster and more accurate technology to use for analyzing the shape and condition of a pipeline.
Compressor Health – new compressors are being manufactured with tightly coupled IoT sensor technology built into the device and provides significant situational awareness to the health status of the compressor.
Acoustic and Harmonics – Detecting sounds and the combination of sounds to derive harmonics is becoming popular in many industries including pipelines. Pipelines ‘sing’ and changes to this song can be indicative of problems and tell operators of issues such as cracks, breaches, and ruptures. The sounds from pipelines can even be subsonic or ultrasonic compared to human hearing but sensors can hear it and understand its consistency. When it varies, this change demands investigation to determine why it varied acoustically.
Proximity – A device that detects the presence or absence of a nearby object, or properties of that object, and converts it into signal which can be easily read by user or a simple electronic instrument without getting in contact with them.
Photoelectric – Photoelectric sensors provide non-contact, accurate detection of targets. They emit infrared, red or laser light and switch to an output state if a target interrupts the emitted light. These units come in three primary types: through-beam, retro-reflective and diffuse. These sensors can check for presence, color, distance, size, shape, and many more applications. They perform at longer distances than alternative sensing methods, while offering many mounting options and the ultimate in flexibility.
Laser – Laser distance sensors are used for precise distance measurement, down to the micrometer at diverse ranges.
Movement – Movement sensors create new possibilities for non-contact detection and the measurement of moving surfaces. The optical movement sensor is compact and designed to detect when surfaces are moving or not. It is ideal for both drive shaft and conveyor applications. The optical movement sensor combines movement sensor functionality with advanced surface scanning technology, to provide speed and length measurements in both x and y direction via incremental outputs.
Contact – Contact sensors are developed to detect the impact of parts as small as 0.3 g.
Magnetic Field – Magnetic field sensors facilitate the safe detection of high electronic current. A typical application would be monitoring pipeline grounding problems.
Radar – Radar sensors are used in different applications for non-contact detection of objects. This mode of operation allows for concealed mounting behind a protective wall, enabling the detection of any movement through the wall. These radar sensors distinguish between approaching objects and objects moving away, up to a distance of 6 meters. The sensors can be adjusted to mask small objects in favor of large ones.
Ultrasonic – Ultrasonic sensors are used in automation tasks for distance measurement and as proximity switches. The sensors use sound/time measurement that ensure reliable detection, regardless of color. Transparent objects, liquids and powders can also be reliably detected. These sensors are insensitive to dirt, so they can be applied in environments where optical sensors would fail.
Internal Sensors – Smart PIG – Depending on the model, smart pigs detect cracks and weld defects through magnetic flux leakage or shear wave ultrasound, mechanically measure the roundness of the pipe to detect crushing, or measure pipe wall thickness and metal loss through compression wave ultrasound.
Pipelines move gas and oil from the pumping source to the refinery and downstream to the markets. A great deal of negative press has been seen in the last few years regarding breaks in pipelines and the resulting environmental damage. Many special interest groups strongly oppose pipelines. Yet, they represent the safest, most efficient means to move product downstream to the markets.
These disasters can be averted with smarter technologies deployed to monitor the pipelines and to provide situational awareness back to the operators. Through the Internet of Things, combined with new internetworking technologies, and watched with artificial intelligence empowered to act and react to emergencies, damage can be caught in advance with new preventive maintenance approaches, and remote control and isolation of leakage with 24/7 ever-vigilant cognitive computing platforms to close values and stem the flow of product.
The issue today with pipelines is the dependence on older SCADA solutions widely dispersed over broad spans of the pipeline. We simply do not have the ability to see every meter of the pipeline with these older systems. We need a far more granular solution that is real-time, and not only monitors the pipeline, but has command and control capabilities to manage it.
Artificial Intelligence will play a major role in this next generation pipeline network management system. It is not meant to replace workers but to augment them and collaborate with them to keep the environment safe and protect our natural resources. With a comprehensive end-to-end sensor network tied to analytic engines and to AI. any pipeline breaks can be minimized and contained in seconds instead of hours or days. This high quality solution and rapid response approach will stem the flows and shut down leaks almost instantly. In fact, these systems can precisely predict the point of a potential future leak and through proactive measures, stop leaks before they ever happen.
IoT sensors can now do things that we could never do before. The variety of sensors and the new low cost model make it affordable to deploy on a more granular scale than SCADA could ever be configured. These IoT sensors can deliver numerous data measurements including smart sensors that collaborate and share data all along the pipeline. External data, such as weather information can be gathered and shared upstream too.
Next generations IoT networks based upon LoRa star architectures can cover long distances, as much as 10 to 20 kilometres in each direction from the gateway depending upon the height of the gateway, terrain, and trees and foliage obstructions. Line of sight to the nodes is important. New mesh architectures can offer higher data rates in a relay topology that permits bidirectional hops along the pipeline and can overcome many issues related to line of sight and terrain as they bend and twist in concert with the pipeline. Both internal and external sensors can be used on the pipelines.
Back-hauling the signals from the pipelines with new satellite connections can also make the links robust and low cost. Reaching gateways in rural or remote locations where no conventional back-haul technology exists can be solved with satellite links. New satellite solutions that offer very low cost connections are now available that make use of BGAN (Broadband Gateway Access Networks) satellites or LEO (low earth orbit) transit satellites that operate in arrays.
The new IoT technologies use minimal power so the sun can energize solar panels to charge up batteries and operate for years without intervention. Gateways can make use of modern energy harvesting techniques to power large points of presence along the pipeline. Nearby streams can cleanly generate electricity. Solar panels and small scale wind turbines can be used as well to generate electricity in an environmental friendly manner.
“Wireless sensors are emerging as one of the strongest options for pipeline monitoring applications,” Varun Babu, Frost & Sullivan TechVision research analyst, said in a statement. “With the adoption of wireless sensor network (WSN) technology, on-board computational sensing and wireless communication capabilities, the quality of monitoring will significantly improve”.
“The WSN sensor nodes and algorithms can provide rich information for detection, location and assessment of structural damage caused by severe loading events and progressive environmental deterioration. WSNs can also monitor more data points and be reconfigured more easily than wired sensors,” Babu said.
Edge computing can add to the robustness of the solution with Edge nodes at stations to act independently if the back-haul network goes down. So, the network can be segmented into sections and then these sections can be further subdivided leveraging a ‘divide and conquer’ method to enhance the effectiveness of the connections. All of these nodes can be programmed to operate autonomously and thereby overcome disconnects from the network. Once the networks are returned to service, collected remote data can be refreshed to the cloud.
When combined with other technologies, such as drones, aircraft, ultrasound, harmonic analysis, LiDAR, intermediate sensing, PIG sensing, and even the existing SCADA systems, all of the data sources can be combined to concatenate these disparate readings into one harmonized over-watch solution whereby the situational awareness of the pipeline can be dramatically enhanced. The diversity of data solutions can cross-conform and augment each other to tell a richer story of the health and performance of the pipeline.
The IoT sensors alone are not enough. The data needs to be analyzed and understood by artificial intelligence systems like IBM’s Watson IoT Platform. One major industry challenge is making data available to decision-makers in a way that benefits everyone. So, derived data outcomes need to be shared with the people controlling the pipeline, but also with the managers and field workers operating the pipeline. It is this return path of data that brings the pipeline to life and provides the real-time feedback, analysis, and dashboard parameters so the team can act in concert to the autonomous systems. This same data can be searched for patterns, trends, and history, as well as combined with legacy data and external data. The story of the pipeline becomes vivid and more telling when aggregating data from so many sources.
The speed at which data becomes available and depreciates, its sources, types and trustworthiness all play a critical role in reaching informed, accurate decisions. Although the amount of data available may be a looming “pot of gold,” more data is not necessarily better. Data volume alone is not a sufficient prerequisite for excellence in decision management in the pipeline industry. To contextualize, organize, and draw true meaning from data, operators are turning to artificial intelligence (AI) and cognitive computing to augment the capabilities of their business experts.
Gathering information and curating data are two keys to effective decision management. However, the ability to operationalize data for decision making is the crucial capability operators need to survive in a world of heightened competition and changing risks.
The pipeline industry has always had to balance seeking additional data while not overspending on the operations. This dichotomy has created an environment where information has been challenging to obtain and made risks more difficult to assess. As a result, operator decisions are typically made as much from experience and intuition (art) as they are from empirical evidence (science).
The “art” of pipeline decision making comes from using learned associations and experience to account for missing data:
- For operators, using prior experience to account for unknown risk details has been central to decision making.
- Analyst staff have been conducting reviews of sampled data, developing correlating hypotheses and making decisions by analyzing several years of service disruption loss outcomes.
- When settling claims, operators predict likely outcomes based upon their experience, which is employed to reduce claim costs and avoid unexpectedly large settlement amounts.
With AI, better models can result to forecast product leakage to the litre and therefore predict the impact to the environment and the cost for clean-up.
AI can instantly and autonomously command the pipeline to shut-down at much more granular points to stop any leakage, then notify the company of this action, the actions taken to manage it, and recommend a course of action towards a speedy resolution of the issue. With in-depth, real-time analysis of the pipeline, the AI systems can get ahead of leaks and help get crews assigned to remediate and fortify the pipeline once issues are known – even before they happen. It is this predictive capability that saves money and reduces or eliminates financial impacts to the business.
AI systems like IBM’s Watson IoT Platform has the tools, people, and processes to help operators move from a reactive lightly aware state of operations to a proactive, deeply aware condition.
At the end of the day, it is a win – win scenario. Companies can operate more efficiently, costs are controlled and predictable, damage and lost is eliminated or minimized, and everyone lives in a safer cleaner environment protecting our natural resources all the while enjoying the benefits of clean, safe, low cost energy to heat our homes, fuel our cars, and make our world work.