Unlocking the Power of IoT and Data Integration to Boost Transformation in the Energy Industry Webinar Highlights (2020)

IoT and Data Integration. Java Philippines. Java.

Unlocking the Power of IoT and Data Integration to Boost Transformation in the Energy Industry Webinar Highlights (2020)

Unlocking the Power of IoT and Data Integration to Boost Transformation in the Energy Industry Webinar Highlights (2020) 768 487 Exist Software Labs

Speaker: Mr. Chris Silerio | VP for Operations

Exist is a software development company that has been in the industry for almost 20 years. Five years ago, we started collaborating and developing systems with companies in the power sector, specifically, generators, distributors, electric cooperatives, and others. We are also one of the technology partners of PEMC and IEMOP. For almost five years, we’ve developed IEMOP’s Central Registration and Settlement System (CRSS) as well as their Trading Operations Central Management Systems. We’ve also recently developed PEMC’s Philippine Renewable Energy Market System (PREMS) through the contract of UNDP and DOE.

We are not an IoT solutions company, but whatever we’re discussing here is based on our vast experience in developing various enterprise system projects in different verticals. That was also made possible by the thorough RnDs and TOCs conducted by our technical team. We are also currently collaborating and partnering with hardware and other software companies that offer IoT devices and solutions.


    1. What is an IoT platform
    2. Discuss the necessary components or services IoT
    3. What are the various ways of implementing an IoT platform
    4. How it could coexist and integrate into a data integration (Data Warehouse/Big Data) platform
    5. How to jump-start your IoT integration journey

What is IoT?

It is the internet connectivity of small devices such as sensors, actuators, smart appliances so they could share and gather information to and from each other.

Some of you might be asking, what’s the big fuss? We already have the internet, and most of you in the power sector already implemented SCADA and PLC before. So how is it any different?

IoT addresses a lot of shortcomings of SCADA and PLC. That is a new level of connectivity between devices. The main difference is when any components within the environment are interconnected, it can act on its own without any human intervention. With all the information it has, it provides rich insight into its entire surrounding environment and instantaneously reacts efficiently by making automatic changes and actions. It can also improve performance and even predict failures, just like Tesla cars or autonomous machines that are in manufacturing plants.

To fully appreciate all the benefits of the internet of things in the energy sector, all of its major components, such as generators, grids, distributors, and consumers, should go smart soon as smart grids cannot function without ‘smart’ consumers.

What is an IoT Platform?

An IoT platform is an integrated service that connects devices and generates its ecosystem online. In the simplest terms, an IoT platform is a group of systems and services where you could operate your entire IoT landscape.

IoT Platform Types, Services, and Offerings

End-to-end IoT Platform

Some companies offer end-to-end solutions. They sell everything – hardware, actual sensors, services, software, and connectivity, plus the device management tools. It is a one-stop-shop, but this tends to be a lot more expensive. You could end up buying features or components that might not apply to your current needs, and yet, you are paying for the entire cost. One of the downsides is that it could become a case of classic single vendor lock-in that most companies steer away from.

Connectivity Management Platform

It is an inherent part of the IoT technology stack. It is the one responsible for connecting and converting data to and from devices and IoT core platforms, regardless if it is on-premise or on-cloud. The primary use case for this type of platform is when you partner with a mobile network to handle communications for security purposes. It means that the data would have to travel through a cellular network rather than flow directly to the internet. But of course, it also introduces new challenges and complexities. So you need to establish your business requirements first.

IoT Edge Platform

We see here the edge devices and the communication gateway. In this type of platform, they provide actual devices and software or tools to manage and configure the devices remotely. With this type, you don’t necessarily need to connect to the cloud as these edge devices are, in a way, functioning already as mini computing and storage machines.

IoT Cloud Platform

It is the most common computing strategy. In this setup, all information from IoT devices could be uploaded to the cloud and a centralized repository. It then runs insights with the aid of full-blown machine learning and artificial intelligence tools. Such simplifies the management of device configurations, managing, and monitoring. Whatever you deploy into the cloud, you could also have on-premise. Your cloud could also push data to the on-premise platform or vice versa.

Edge Computing

An essential part when deciding what type of platform you will choose is whether you would go with edge computing, cloud computing, or hybrid.

Edge computing is bringing the processing at the edge of the local network, not the office LAN, but the network within the autonomous component, for example, a smart car and smart parking garage. These edge devices could send and receive from various IoT devices and communicate directly with other edge devices. 

A case in point is the autonomous Tesla cars that can park itself in an IoT enabled parking garage. The cars going in and out of the garage did not wait for details from the cloud. Instead, they talk to each other directly. Such solved a crucial problem associated with the centralization of data in the cloud. In this model, all processing is distributed, and decision making is down locally by bringing cloud computing capabilities and localizing the data to these devices. It can reduce latency, process data faster, and provide instantaneous insights, as compared to sending everything first to the cloud for analysis.

How Can It Do That?

All the IoT cloud components have been, in a way, miniaturized, streamlined, and deployed in those small edge devices.

Common Components of Edge Devices

Complex Event Processing (CEP) could take data from different IoT devices and act accordingly based on the established patterns. But more often than not, actual CEP models are done on an edge core network. Then, it’s just pushed to the edge devices.

Edge computing devices now support Machine Learning & Artificial Intelligence (ML & AI) locally, but there is always the option of sending the data to the enterprise data warehouse for furth ML & AI processing.

It has the ability to deploy containerized applications as well, either the software package from your solution provider or your own developed applications.

These devices are capable of storage and data management too. These devices now have storage large enough to store all data from different devices, especially in the event of network failure. It is capable of holding data temporarily until the connection gets restored.

When do we consider EDGE over CLOUD?

When your requirements are mission-critical, like, when a machine needs to stop outright to prevent an accident or defective product, sending your data first to the cloud for analysis is not an optimal solution.

Another scenario is when these components are remote, and there is limited or no connectivity to your centralized data center.

Edge computing is still developing, so it has a lot of downsides. Given that the data is localized, security measures might not be as tight as the one securing the cloud security-based systems. It is still a potential privacy and security concern in the edge devices given that the data is stored and already consolidated. That information could also contain sensitive data. Curious people can easily access the device, have the data leaked, or used maliciously.

High availability and failover. Usually, these devices are geographically dispersed. If one edge device goes down, even if you have installed clustered edge devices in one autonomous component, it is still hard to monitor and replace if you have a lot in your IoT plinth.

Scalability. As the number of IoT devices grows, it will be hard to maintain and monitor. Even if you did conduct positive planning, it could only handle so much.

So, choosing between edge and cloud computing depends on your business requirements as well as your business objectives.

Data Integration Platform

The data integration platform is another significant type of platform. Whenever you see an IoT platform presentation, there is always a storage component. It is a combination of several frameworks and tools that will handle the following 

    • Streaming of extracted data from various sources 
    • Routing of data either to your data warehouse or directly to the data consumers’ application
    • Ingestion of data to your data warehouse and OLTP, 
    • Doing a fast filling of data from the database. 

All of these are components of a full-blown enterprise data warehouse or Big Data. What we would like to note is that when you are planning to start your IoT platform, you should first have a data warehouse or Big Data strategy in place.

The goal is to align your IoT storage roadmap with your data warehouse or Big Data roadmap, and not the other way around. Remember that a data warehouse or Big Data setup is also a costly investment and should be treated as a separate project. Having an aligned IoT and data integration strategy could save you a lot of money by maximizing the use of your entire infrastructure.

There are a lot of components that IoT and data warehouse share – streaming, ingestion, and routing. All of these can have its cluster but can be utilized by both your IoT and data platform. With this setup, you can save cost in your infrastructure and reduce the skillset that you need from your IT team.

IoT and Big Data share the same challenges – volume/size of the data coming in, velocity/speed, variety, and veracity/accuracy. When deciding on your IoT storage platform, you have to make sure that it could address all of these challenges.

Here are the few options that you could opt to adopt for your IoT and data warehouse platform.

So, when you have to decide as to which computing strategy or data platform to employ, you need to consider a lot of factors. Don’t just jump on the bandwagon.

Speaker: Mr. Steven Siahetiong | Exist Technical Architect

IoT Components

  1.     Things pertain to sensors, actuators, or any device capable of sending data to the cloud.
  2.     For these devices to send data, it will need to use networking and messaging protocols for the data to be transferred to its destination. Most use wireless access for network connectivity. During the past few years, there have been substantial developments in wireless connectivity protocols. Some of the examples are Bluetooth, Low Power WAN, ZigBee, 6LoWpan, and Thread.
    • Bluetooth is a global 2.4 GHz personal area network for short-range communication. 
    • LPWAN is a type of wireless telecommunication wide area network designed to allow long-range communications at a low bit rate. There are several competing standards and vendors in the LPWAN space. The most prominent of which include Laura and SigFox.
    • ZigBee is a 2.4 GHz mesh local area network lan protocol. It was originally designed for building automation and control, so things like wireless thermostats and lighting systems often use ZigBee.
    • 6LoWpan uses a lightweight IP based communication to travel over lower data rate networks. It is an open IoT network protocol like ZigBee and is primarily used for home and building automation.
    • Thread is an open standard built on IPV6 and 6LoWpan protocols. You could think of it as Google’s version of ZigBee.

For the messaging layer, the most popular are HTTP, MQTT, and CoAP.

    • The CoAp protocol is a client server-based protocol that allows pockets to be shared between client nodes, which are commanded by the CoAP server.
    • The MQTT protocol is communication-based, which is based on the publish/subscribe methodology in which clients receive information through a broker only to the subscribed topic.
  1.     The IoT solution needs to have a platform. An IoT platform combines several IoT functions in one. It can collect and receive your data, convert data between protocols, and store and analyze data. They are available as cloud-based and standalone platforms and are available for many companies, both large and small.

The cloud will have an important role to play in IoT as it will enable companies to create networks, store data, and automate processes without having to build the infrastructure themselves. Such will let IoT services to be developed much quicker and at a lower cost than using traditional in-house systems and services.

As seen on the graph, HTTP involves the largest bandwidth and latency than any other protocols, while CoAP has the least bandwidth and latency.

MQTT offers the highest level of quality of services, with the least interoperability among the four. On the other hand, HTTP was designed for the greatest interoperability on the web and did not include reliability as a core feature.

There are many IoT platforms in the market, and the functionality of these platforms varies enormously. Although all IoT platforms will have dashboards to display data, some platforms are essentially dashboards and are only capable of displaying data from devices. You will often find the terms dashboard and platform used interchangeably.

An IoT dashboard can be considered as a basic IoT platform. A dashboard can usually display data and control devices. However, an IoT platform can usually collect data from various sources, store data, control devices, display data, run tests, deploy device updates, and manage device inventory.


Events, insights, and actions are functional concepts that exist across the devices, platforms, and applications of an IoT solution. To further explain, consider an application that monitors the cooling system, temperature for food storage, and calls emergency maintenance services if the temperature becomes dangerously low or high.

The following processes occur in this example:

    • The devices send temperature samples from the primary cooling system to the IoT gateway via the device to cloud events every 30 seconds.
    • These events can generate insights. The IoT platform can evaluate events for any immediate contextual insights, such as temperatures at malfunctioning levels.
    • The generated insights can trigger actions. If the temperature is at a malfunctioning level, the platform can send a command to the backup system to start while the maintenance is en route to the location.

Events represent the device-to-cloud communication in an IoT solution, and maybe notifications, acknowledgments, or telemetry data.

Insights are interpretations of events. It may derive from events directly as contextual insights or transform or stored event data by application event processing for real-time or aggregated insights.

Actions are activities undertaken either programmatically or manually as a device, service, or analog actions.

IoT Customer Scenarios

How can you design an IoT solution to satisfy customer requirements?

    • Manufacturing safety systems must respond to operational data with ultra-low latency and control.
    • Mission-critical systems such as remote mining equipment, connected vessels, or offshore drilling need to analyze and react to data even with limited connectivity to the cloud.
    • The volume of data produced by jet engines or connected cars can be so large that data must be filtered or pre-processed before sending it to the cloud.
    • Regulatory compliance may require some data to be locally anonymized or aggregated before being sent to the cloud.

IoT Edge

IoT devices can connect to the IoT platform directly or through IoT Edge gateways that implement intelligent capabilities. With an IoT edge, you can analyze censored data in near real-time and issue commands when you detect anomalies to stop a machine or trigger alerts. Your streaming logic runs independently off the network connectivity, and you can choose what to send to the cloud for processing or storage. You can also filter or aggregate the data that needs to be sent to the cloud.

The benefits of using edge gateway, particularly in an IoT application, is that moving data processing functions from the cloud to the edge helps ensure accuracy and reliability. Transmitting data to and from the cloud takes time. Even milliseconds are too long for many mission-critical decisions and processes that occur in industrial operations. In the time it takes to send data to and get a response to the cloud, the data could simply become obsolete, resulting in missed opportunities for action, damaged components/products, or risks to the safety of the equipment/personnel.

Edge devices can also send data or commands directly to other devices for immediate action. Edge computing via edge gateway also makes it possible to keep sensitive data on-site to ensure its security. Some devices or systems generate so much data that the bandwidth and resources needed for the cloud to handle this data are too costly. The edge gateway can determine which data can be sent to the cloud and transmit it in the most efficient and usable form.

Industrial Edge LORA 1 Gateway

    • Extremely flexible and powerful
    • Allows to run Lola 1 application and network server locally to setup on-premise private network
    • Aimed to help the transition from legacy to connected automation with its ability of interfacing existing devices.

Types of Edge Gateways

IoT Edge devices can act as communication enablers, local device control systems, and data processors for the IoT cloud platform. IoT Edge devices can run cloud workflows on-premises and can communicate with devices even in offline scenarios.

Cloud gateways can do protocol and identity translation to and from the IoT cloud platform and can execute additional logic on behalf of devices.

In a transparent gateway pattern, the gateway simply passes communications between the devices and the IoT cloud platform.

A protocol translation gateway is also known as an opaque gateway, in contrast with the transparent gateway pattern. In this pattern, devices that do not support MQTT, AMQP, or HTTP can use a gateway device to send data to IoT cloud platforms on their behalf. All information looks like it is coming from one device, the gateway. 

Exist, as a technological solution provider, designed a generic IoT and data integration reference architecture for our customers, especially for the power industry.

The Data Integration layer handles the ingestion of the various data sources. It supports all types of connectors.

IoT Gateway acts as a gateway to connect your devices and accepts data streams using MQTT, CoAP, and HTTP messaging protocols. Data can be ingested as a batch or a real-time stream. With batch processing, data is collected in batches and then fed into an analytics system. A batch is a group of data points collected within a given period. Unlike stream processing, it does not immediately feed data into an analytics system, so results are not available in real-time.

With stream processing, data is fed into an analytics system piece by piece as soon as it is generated. Instead of processing a batch of data overtime, stream processing feeds each data point or a micro-batch directly into an analytics platform. This allows teams to produce key insights in near real-time.

Streaming platforms allow data transformation persistence and allow interactive queries on the streaming application. Kafka is the most popular choice among streaming platforms. 

ETL is the extract, transform, and load process for the data. While the ELT is the extract, load, and transform process. In ETL, data moves from the data source to staging and into the data warehouse. ELT lets the data destination to the transformation, eliminating the need for data staging.

Optionally, you can add a dataflow tool such as NiFi to provide a web UI, which design or control your data pipeline in a graphical representation. NiFi can deal with a great variety of data sources and formats.

For Stream Analytics, the popular choice is Apache Spark Streaming. It is a scalable and fault-tolerant stream processing system that allows data engineers and data scientists to process real-time data from various sources.

Data can be processed using complex algorithms, expressed with high-level functions. Finally, processed data can be pushed out to file systems, databases, and live dashboards. You can apply Spark’s machine learning and graph processing algorithms to data streams.

For the ETL jobs, you can apply data quality steps before loading it to a data warehouse. While for ELT, data is simply dumped into the data lake, and transformations happen on an as-needed basis, and only the data that needs to be analyzed at that time are transformed.

Hot path

    • analyzes data in near-real-time, as it arrives
    • events must be processed with very low latency
    • typically implemented using a stream processing engine
    • the output may trigger an alert or be written in a structured format that can be queried using analytical tools

Warm path

    • holds data that must be available immediately from the device for reporting and visualization

Cold path

    • performs batch processing at longer intervals (hourly or daily)
    • typically operates over large volumes of data, but the results don’t need to be as timely as the hot path.
    • captured and then fed into a batch process

Machine Learning allows predictive algorithms to be executed over historical telemetry data, enabling scenarios such as predictive maintenance. It requires training by telling a machine learning model what it’s trying to predict, similar to how a human child learns.

Speaker: Mr. Jejomar Dimayuga | Exist Technical Architect

Industrial IoT Stack


    • Actual things that are located in your physical environment
    • Smart meters or actuators

Edge Devices

    • Can act as the concentrator/aggregator for all of your devices
    • Allow different protocols depending on the support of your devices


    • Connectivity to your edge devices

Cloud (or on-premise infrastructure)

    • Where your applications run
    • Offers high availability and high scalability solutions


    • Where analytics are produced that can be based on historical events, failures, rules, or condition that can communicate back to your device
    • Where algorithms are performed after ingesting and storing the data


    • How your data is consumed
    • How can it be displayed via the web, iOS, or Android devices

Security and encryption across the components are must-have. It is a vital requirement for all systems, especially in the IoT solution for energy. Security components such as SSL or HTTPS, LDAP Integration, Single Sign-On, Access Token are already in place.

We accommodate Access Token so we can ensure that your devices are secured in connection to your edge devices and IoT gateway.

We also offer a Single Sign-On authentication to your existing authentication features via the Oauth2.0 protocol.

We can also integrate your authentication via LDAP. We can connect your authentication and your road-based access control to an active directory in your organization if you have one.

To secure your communication layer, we can incorporate SSL or HTTPS with certificates from the latest update of the TLS version.

Another important aspect of any system nowadays is scalability and high availability. In this solution, we can incorporate Docker containerization via the Kubernetes platform. In this way, we can deploy your application or your IoT gateway into multiple replicas to ensure that it is highly available. We can:

    • Reach the maximum daily quota of your required messages
    • Reach the quota of connected devices 
    • Increase the ingestion throughput
    • Increase the processing throughput
    • Set up a DR environment in place to ensure that your operations will be continuous

All of these are in place already so that we can ensure that you will be able to perform your analytics and your business needs. On top of that, we can also integrate into your existing systems, may it be your billing system, billing, or offer system, CRSS, or else.

Speaker: Mr. Chris Silerio | VP for Operations

Things to Consider on Choosing an IoT Platform

    • The capability of your team
    • Geographic coverage
    • Timeline and budget
    • Your IoT and EDW roadmap
    • Your business billing model
    • Scalability of the platform
    • Method of connectivity
    • Integration flexibility
    • Device management
    • Cloud VS. on-premise

Developing your platform in-house is fine as long as you are not pressed by time. Otherwise, it might be better to start with a simple platform and develop on top of it.


Q1: With the vast array of open standards for all the layers of IoT, is there a considered safe or conservative stop that anyone can use as a basis or a starter that is easy to pivot towards specific solutions as requirements get clearer? 

A: As shown on the Industrial IoT Stack diagram, we could have the base layer of the devices, followed by the edge devices if you want to go with edge computing. After that, we can set up an IoT gateway on cloud infrastructure.

It also depends on the network or protocol that the sensors, actuators, or your device are supporting. What matters most is that your platform can connect with these various open standards and not limit your entire solution. If there is a new protocol that comes in, your platform should be flexible enough to support these open standards.


Q2: What are actuators?

A: Actuators are mostly used in factories. These are devices that are connected to types of machinery. It typically sends telemetry data, like temperature, weight, the roughness of a substance.


Q3: What is the experience of Exist in developing IoT systems for energy efficiency applications?

A: We have a couple of projects on the integration of IoT devices, and it is still growing. Our forte is software development integration. But our experiences with these projects gave us enough exposure to the entire ecosystem.


Q4: Have you done any IoT applications for use in the Philippine market?

A: We are developing some right now for companies who are just starting their IoT projects. These companies are doing it in phases. We are currently in the first phase. (Company names cannot be disclosed yet.)


Q5: Given the poor internet connection in the Philippines, would IoT applications that HTTP use be a challenge?

A: High likely, cloud computing would be a challenge because it will always depend on internet connectivity. In scenarios that it is in remote areas, it would be better to go with edge computing provisioning. That is one of the advantages of an edge over cloud computing. 

Aside from internet connectivity, some providers also offer alternative communication layers, such as cellular networks.


Q6: Do you provide a complete solution, like sourcing sensors or edge devices, depending on the requirements?

A: No. There are only a few companies here in the Philippines that offer end-to-end solutions. We are not a hardware company. We are agnostic of software and hardware. We do more on the development of all these layers/components from the IoT devices up to the data integration. But we are partnered with companies that we may utilize to provide an almost end-to-end solution.


Q7: For an organization that has mostly in SCADA plus TLC equipment, what would a migration plan look like when moving to a more modern platform that allows a more open choice of technology stack?

A: IoT is not intended to replace SCADA and TLC. It is a different level/layer. Information from SCADA and TLC may be integrated into your IoT platform. We do not have to migrate as we can support SCADA and TLC.


Q8: With regards to data storage, would you still recommend Hadoop, or are there better alternatives?

A: We recommend that you go with the storage that can handle multiple parallel processing and the other factors that would support volume processing of data, and speed from streaming data, flowing data from different data sources to your storage devices. There are a lot of databases that support multiple parallel processing, like Greenplum, Oracle, Teradata, and Hadoop.


Q9: If I were to engage for an IoT application, what would be the start?

A: It might be better if you could consult first with a third-party company that already has the experience or an offering of IoT solutions in the market. It is quite hard to say outright what the specific software that we are going to use in your IoT platform. We need to undergo some discussions within your organization before we could provide a straightforward platform that you could use. We could always start simple. But regardless of how simple it is, we still need to know what are the factors that we should consider for your organization.


Q10: Can you offer an outsource service to support an IoT application?

A: We could somehow handle it, given that we have partners. You may also directly talk to hardware vendors, especially about edge computing devices. We can also give you a list of vendors that you could contact for each IoT component