Understanding the IoT value chain. Where does an IoT product fit?

Last update on 2025, 14th of Jan.

Copyright

Copyright 2024 SpimeSenseLabs . All rights are reserved. This article and its content are protected under copyright law. No part of this article may be reproduced, distributed, or transmitted in any form or by any means, including photocopying, recording, or other digital, electronic or mechanical methods, without the prior written permission and mention of the copyright owner, except in the case of brief quotations embodied in critical reviews and certain other noncommercial uses permitted by copyright law and only by mentioning and referring to the copyright owner. For permissions requests or inquiries, please contact SpimeSenseLabs . [email protected]

Introduction:

In this, live on going, article I'm explaining the whole IoT value chain building blocks and its component in addition to the value for each component to spread the knowledge among my colleagues and students. In every part/category of the value chain I'm also listing some specific providers and companies that had an important role in this category over the last 10 years even if some are not in the IoT business today whether due to successful acquisition, failure or went out of business for a reason or another. Please leave me a comment if you believe your company should be mentioned in one of the IoT value chain building blocks and i'll keep updating the article from time to time to keep it up to date.

Lets start by basic understating of the building blocks of internet of things and the value of each block.

It is easier to understand the IoT solutions value chain (building blocks) by comparing it to the mobile solutions value chain. The provided figure in the top of this article illustrates the path of the mobile application data traffic in the upper part and compares it to the IoT data traffic path in the lower part.

In the mobile world, the building blocks start by the mobile phone (the end user device) and it runs an application client (i.e. WhatsApp, Facebook, etc.) on top of the phone operating system (i.e. Android, iOS, etc.). When the mobile is being used at home, the mobile sends the data through its local Wi-Fi network to the Wi-Fi access point/router at home (the Wi-Fi Gateway) and the router transfers the data to the internet through a carrier network (ADSL copper cables, fiber cables, 3G/4G/5G network or satellite network) that carry the traffic though a very long distance (sometimes hundreds of kilometers) to its final destination application server on the internet (cloud server on the internet) .

Similarly, the IoT components and building blocks start by IoT Sensors/Devices (computerized device similar to mobile phone) whether it is a smart meter, street lighting pool, digital signage screen or parking sensor.

The sensor/device data is transmitted through a small distance wireless network called wireless sensor networks (somehow similar to Wi-Fi network at home) to reach the IoT Gateway (very similar to Wi-Fi Access Point). The IoT Gateway transfers the data to one of the carrier networks that carry the data through a long distance to its final destination on the Application Enablement Platform (compared to application servers) that runs a server side application (compared to web application).

Each of those components and building blocks is explained in more details in later sections. if you prefer to learn it through online tutorial videos, you can find it in great online free tutorial available in the link below.

https://maharatech.gov.eg/course/index.php?categoryid=8

Now let’s dive into the value and purpose of each block in the IoT value chain and which company is doing what in each block and its logical value.

Let’s dive in more details of each component (building block) of the value chain and the variants or sub-blocks in each component/block.

The value chain consists of 8 building blocks/components that are marked in the image from 1 to 8 and each of these components have sub-components grouped in categories with multiple alternatives for each group/category. The value of each of these alternatives is explained below with samples of the various alternatives in each sub-category.

1- Sensors/Devices:

These are the small computer devices usually called sensors and sometimes called objects that sense something to send it to the outside world.

Some device maker companies like Libelium build several IoT devices and have a large portfolio of devices. Other device makers like Eurotech , are smaller and less active. Other device makers might be focused on specific devices like @Diehl Metering who are focused on smart metering companies . Most device makers can customize a device for a special use case but the ones who can do it right are few. When a device maker starts creating a device, as a quick way to have something working to show it to the stockholder or customer, device makers usually start the device development/creation phase by utilizing a prototyping board like Arduino, Intel Edison or any of the commercially available hundreds of prototyping/development boards or even mini/small computers like Raspberry Pi.

If that prototype is good, the device maker through it all away and start building the real actual/final device by choosing the right low power wireless technology modem chip, the right MCU chip, enough memory chip, screen, buttons, sensing elements/modules and other components required for the device and fix all of these components together on custom PCB (Printed Circuit Board) and package all of that inside a suitable enclosure/box.

These small devices have software that is running on top of an operating system inside the device. This device software is usually called device firmware or embedded system software. Some companies like Witekio are specialized in writing/developing the embedded device firmware only and not making the whole device.

Depending on the complexity of the functionality required in the device, the device maker will choose a suitable RTOS (Real Time Operating System) for their device.

RTOS (Real Time Operating System):

Similar to a mobile phone, a lot of IoT Devices need to have RTOS. Some of the IoT devices are doing multiple functions each function is separate in a software application and hence these devices are running RTOS (Realtime Operating System) comparable to the Android OS on your phone and in this case the device RTOS will be running a group of applications on top of it like the applications installed on top of your Android phone or applications installed on our iOS if you are Apple phone user. Some IoT devices are doing very basic functionality and hence it is using a firmware that directly interacts with the device HW without relying on an RTOS.

There are lots of Operating systems whether of commercial license or open source. A very interesting open source RTOS list with a brief description of each RTOS and its supported HW platforms/architecture and supported wireless technology are available at OSRTOS.com

The selection of the suitable RTOS depends on the specific requirements of the required embedded system including device resource constraints, real-time capabilities, scalability needs, community support and the intended use case. Contiki-OS and Riot-OS are tailored for IoT devices, while FreeRTOS and Zephyr are more general-purpose, and TinyOS is specialized for wireless sensor networks. OpenWrt, on the other hand, is suitable for bigger devices used as networking equipment (IoT Gateway).

Zephyr and Riot-OS are particularly well-rounded for MQTT-enabled IoT applications, while the others, depending on the time of implementation, may require additional customization or external libraries to support MQTT protocol

Here is a brief comparison between some famous RTOS, thanks to ChatGPT and Gemini

RTOS comparison conclusion:

??????? Open Source:?All listed RTOS (Contiki-OS, Riot-OS ,Zephyr, TinyOS and OpenWrt) except FreeRTOS are fully open source,?offering better customization and community support.

??????? MQTT:?All above RTOS support MQTT directly or through 3rd party libraries,?making them suitable for IoT devices.

??????? Ease of Use:?ContikiOS and TinyOS are easiest for beginners,?while Zephyr and OpenWRT have steeper learning curves.

??????? Main Features:?Consider Zephyr for security and multi-core,?RiotOS for low power,?ContikiOS for IPv6,?and OpenWRT for extensive network features.

RTOS alternatives in more complex IoT devices: Those devices include Robots, smart cars, self driving cars, etc. These devices have a lot of components and subsystems and software modules as well. Makers of these devices tried to build their own specialized RTOS alternative or an RTOS like system inside the device. . The 3?? layers architecture of #Autosar in as middle layer managing all the underlying ECUs (Electronic Control Unit) and managing all of the applications on top ?? of Autosar

2- Wireless Sensor Networks:

Wireless communication to transfer the data through small distance. This distance can be very small up to 1 cm like the case in NFC (Near Field Communication) and can be few meters (up to 10-15 m) in case of RFID (Radio Frequency Identifier) , larger distance of like BLE (Bluetooth Low Energy) where the wireless signal can reach to 300 m in BLE v5 or few kilometers in other wireless technologies like the case of LoRa, LoRaWan and Mioty. All wireless sensor networks are meant to be for low power consumption hence the name LPWAN (Low Power Wide Area Network) where the wireless technology save the device power either by minimizing the data throughput or minimizing the distance to save the device battery. Experts' advice on What is the best way to design an IOT network for a smart city, is available on the link below

https://www.dhirubhai.net/advice/1/what-best-way-design-iot-network-smart-city-ynb8c?utm_source=share&utm_medium=member_android&utm_campaign=share_via

?

3- Gateways:

The gateway is supposed to take the data from a small distance wireless technology (i.e. BLE, LoRa, etc.) and transfer it to a long-distance carrier network (sometimes up to hundreds of kilometers). The better the gateway the more wireless technologies it can support and that is why companies like 英特尔 develop a board with modem chips for multiple wireless technologies and supply it to the gateway makers (i.e. Cisco Networking ) to use it inside their gateways with intel inside logo.

Some companies like Faltcom started building industrial IoT gateways that can accommodate working in tough conditions. later it has been acquired by Telia and now the gateway is branded as Telia IoT Edge that is used as a gateway to connect public buses in Sweden to the internet.

Alleantia also built an industrial IoT gateway that is now branded as industrial edge platform.

Please note that Gateways represent the edge of the network and hence edge applications run on the gateways. A detailed explanation of the edge and why we need edge applications is available in the link below

Let's clear the confusion of #EdgeComputing, why do we need an #Edge Application and where would an Edge #Application run in an #IoT solution?

large networks contain multiple gateways to cover larger areas. Each of the gateways needs to authenticate the device and check if the gateway will allow the device to use the gateway to connect to network or not. Each gateway cannot decide this by itself but rather rely on a higher decision authority known as the network server that take a central decision for all the gateways in the same network when it comes to each device trying connect to the network.

LPWAN Connectivity Management Platforms (Network Servers):

1. The Things Network (TTN): TTN is a community-driven initiative that provides an open-source LoRaWAN network server. It allows users to set up their own private network or join the global public community network. The source code is available on GitHub.
2. ChirpStack: Formerly known as LoRa Server, ChirpStack is an open-source LoRaWAN network server that provides scalability and flexibility. It is written in the Go programming language and is available on GitHub.
3. LoRaServer: LoRaServer is an open-source LoRaWAN network server developed in Go. It supports the LoRaWAN 1.0 and 1.1 specifications and provides a RESTful API for integration. The source code is accessible on GitHub.
4. OpenChirp: OpenChirp is an open-source project that includes a LoRaWAN network server. It is designed to be scalable and extensible, allowing for customization. The project is available on GitHub.
5. Loriot: Loriot offers an open-source version of its LoRaWAN network server. It supports various deployment scenarios, including public, private, and hybrid setups. The source code can be found on their GitHub repository.
6. MioTy Service Center: The Network Server in MioTy terminology is called/named a Service Center. Unfortunately, it is a confusing name appeared in MioTy documentation dated 22.04.2021 and still being used in their new developer portal on 19 Jun 2024 and hopefully it will be fixed soon. MioTy Network server (Service Center) is responsible for network management tasks in the MioTy Networks including Aggregation of uplink data received by MioTy Gateways (Base Stations), Scheduling of downlink data that will be sent by the Gateways, Distribution of devices' (End Point) information (Device IDs and Device keys) to MioTy Gateways.

Examples of the MioTy Network Servers (Service Centers according to MioTy confusing terminology) are MyThings and ResIoT.

1. MyThings: MyThings is developed by Behrtech. MyThings is a MioTy central network connectivity platform with device management capabilities designed for MIOTY-based low-power wide-area networks. In this context, MyThings is comparable to both Network Server and Application server in LoRaWan. in MioTy terminology, it is a mix of MioTy Service Center and MioTy Application Center (MioTy basic device management server).
2. ResIoT MioTy Cloud: Developed by ResIoT as MioTy Service Center and MioTy Application Center. in MioTy confusing terminology, The Network Server is called MioTy Service Center and the Application Server (basic device management servers) is called Application Center. RestIoT offers a free cloud based version with a limited number of devices (15 devices) and 5 days data retention. In this context, MyThings is comparable to both Network Server and Application server in LoRaWan. in MioTy confusing terminology, ResIoT MioTy Cloud is a mix of MioTy Service Center and MioTy Application Center (MioTy basic device management server).

4- Carrier Networks:

Carrier networks are networks that can carry the data to long distances in several km distance. These networks could be mobile network ?? (2G, 3G, 4G, 5G). These mobile networks are also called TN (Terrestrial Networks) and are operated by mobile CSP (Carrier Service Provider) Like Orange Egypt , Vodafone , Deutsche Telekom in Germany or Verizon Communication in USA. Carrier networks have evolved to suite IoT and the most recent commercially available evolution is the 5G networks. Check the dedicated article What is in 5G for IoT?

The carrier network could be copper cables or fiber cables operated by fixed carrier operator known as fixed CSP like Telecom Egypt or a carrier network could be a LEO (Low Earth Orbit) satellite ??? networks. Satellite networks are also called NTN (Non Terrestrial Networks) and are operated by satellite ??? network providers like Lacuna Space , STARLINK , Eutelsat OneWeb , OQ Technology , Kinéis, Inc. , FOSSA Systems , Hello Space , etc .

Thousands of these LEO satellites are now circulating around the earth broadcasting and receiving from IoT devices and hence, interfering with each other.

More details on the satellite networks are provided in the links below.

• Details about the radio interference caused by LEO satellites are available in this post.
• Satellite ?? connectivity requires ground stations to transmit the data back to internet on the earth, and hence #Satellite ?? #IoT, is gaining more momentum all over the world from all parties including investors on the #ground.
• Very good summary of current situation and future requirements for #IoT in a seamless world of the ?? #TN mobile ?? networks ?? and #satellite ??? #NTN.
• Summary of GEO/LEO satellite ??? business in communication infrastructure and it's expected customer use cases compared to traditional terrestrial 4G/5G network.

Inside the carrier networks, there are lot of sub-categories, sub-systems and sub-topics

4.1 Low power carrier/operator networks:

Sigfox 0G technology started the first low power carrier/operator business by catering for saving device power over data throughput allowing devices to service for 7 years on a battery without recharge (actually battery replacement) and had their networks in 70+ countries as one global network that is acquired later by UnaBiz and became SigFox UnaBiz. Because 3G and 4G networks consume a lot of power from the device battery, the competition from Sigfox pushed the network makers/vendors to look into ways to make their networks consume less power from the device and giving away the high data throughput. 爱立信 proposed NB-IoT/NB LTE, 诺基亚 proposed LTE light and 华为 proposed cellular IoT and at the end 3GPP standardization body adapted Ericsson proposed changes in the 4G network to enable mobile operators/CSP to start offering NB-IoT/NB LTE for IoT devices to have bidirectional (uplink/downlink) low power mobile communication with a better throughput than SigFox and with an IP protocol as well but they still inherit the problems in roaming agreements if the device will be moving between countries where data rates/tariffs are higher in the visiting countries compared to data rate/tariff in the device home network in its home country. 5G networks are coming by default with NB-IoT , mMTC and Passive IoT modes of operation for IoT devices.

4.2 Global MVNOs (Mobile Virtual Network Operators):

Global MVNOs appeared to solve the roaming problem in NB-IoT, some companies like KORE , Arkessa (acquired by Wireless Logic Ltd) started utilizing eSIM capabilities to provide a global sim to eliminate the high roaming data tariffs for traveling devices for the global device makers who would like to transmit data from their devices wherever these devices would in the world. Other companies followed this model include 1NCE , emnify, Onomondo, Velos IoT, Eseye, floLIVE FloLive, Shield IoT, etc.

4.2.1 eSIM:

SIM (Subscriber Identity Module) is a physical electronic chip packaged in small plastic cover. this chip holds what is called IMSI (International Mobile Subscriber Identity) number stored inside the SIM. The SIM is provided by mobile network operators to the subscriber to be placed inside mobile subscriber's phone or IoT device for the network to identify/recognize the subscription of this phone or devices and hence these phones or devices get connected to the network for the sake of getting internet connection as well through the mobile network.

Over the time, some phone manufactures manufactured phones (or devices) that could have dual or triple SIMs to have access to dual or triple mobile networks. the SIM size (Form Factor) has changed over the time to smaller in size and hence called mini SIM and got more smaller and hence called Micro SIM and then came the idea of an eSIM which is a software version instead of the original electronic chip. Think of the eSIM as small file that contains the IMSI number and it can be remotely downloaded/sent/pushed to the phone or IoT device. The beauty about eSIM that it became possible for phones or IoT devices to multiple values of these eSIMs and in this case each value is called an eSIM profile. For more information on what is eSIM and how it is being used, check the LinkedIn post in this link to download a good eSIM guide developed by GSMA - Internet of Things

4.3 Connectivity Management Platforms:

Mobile operators (weather Global MVNO or local operator) figured out that they would need to have an IoT CMP (Connectivity Management Platform). CMP is a software that make it easier to manage several connectivity subscriptions (SIM cards) for a specific customer (usually a corporate) where the customer would need to enable/disable the subscription or monitor the amount of data generated from every SIM (representing a device). The first company that built a connectivity management platform was called Jasper wireless that have been acquired later by Cisco and later the product is branded Cisco Connect and now later branded as Cisco Control Center. Mobile network providers/vendors (like Ericsson, Nokia and Huawei) also started building their own CMPs. Ericsson made Ericsson DCP with some billing capabilities as well. later Ericsson branded it as Ericsson Accelerator and later sold it to Aeris . In 2017 Nokia started building its CMP and called it Nokia WING (Worldwide IoT Network Grid) even that the name implies a global network but it is more of a CMP and looks like Nokia was dreaming of utilizing CMP with eSIM but it never happened and ended up Nokia not mentioning Nokia WING product anymore. Some operators also built their own CMP like Verizon in USA who built their own CMP and branded it as ThingSpace and similarly Vodafone built their own CMP and called it Vodafone DCP (Device Connectivity Platform) and later rebranded it as Vodafone Business IoT and in the begging of 2024 Vodafone moved this CMP business out of Vodafone in a new separate joint venture company between 微软 and Vodafone.

More about the CMP is available in this post

5- IoT AEP (Application Enablement Platform):

The IoT AEP is the smartest and most complex component in the whole IoT value chain where all the IoT magic happens. Few people (non-field experts) refer to it as the IoT middle-ware because it is a middle layer between devices and applications. There are a lot of companies claim that they offer an IoT AEP. An IoT AEP must have certain capabilities and offer certain features to be considered application enablement platform. I'll put the AEP capabilities and features in details in a separate dedicated slide/section/article.

In this section, on the sub-sections below, i will divide the IoT AEPs into families based on their specialization, focus or targeted industries or customer segment

5-1 IoT AEP family for boosting Productivity, efficiency and guarantees neat solution Architecture:

This is a family of IoT AEPs that focus on boosting the productivity and efficiency of IoT application development process. This family includes MasterOfThings horizontal IoT AEP for those would need to develop various IoT use cases/solutions like telecom operators and smart cities while for IIoT (Industrial IoT) ThingWorx, a PTC Technology is a good example in this category. Platforms in this category are all licensed software whose main purpose is to empower the solution owner (i.e. smart city management or telecom operator) to have control over its destiny of operation and have suppliers independency (device agnostic) and application provider agnostic and very much focus on empowering the developers to do more applications quickly, efficiently and properly usually though providing 4 main core modules of the IoT Platform. Check this post for a quick/brief of how a platform empowers developers.

The 4 main modules of a professional IoT AEP (Application Enablement Platform) are

A- Visual IDE (Integrated Development Environment): a visual IDE is very important to faster and facilitate the development of the use cases/applications' backend services as well frontend end user interfaces weather it is web or mobile (Android/iOS). This module is very hard to develop and, hence, very few platform providers provide it as a core module of their IoT AEP . All the weak platforms leave it for the developers to rely on traditional development tools and provide the developers with a big list of platform APIs to use where developers have to live with lots of integration work and logic outside the use case/application logic.

B-Sensor profile definition: a sensor profile definition module with an easy to use GUI is important to define device sensing capabilities and information about the device profile (i.e. readings, reading type, encryption key, its installed location, etc.).

C-Data collection/acquisition, monitoring, enrichment and orchestration module: a data collection module is required to collect data from the devices and execute real-time actions (events) upon data arrival usually relying on a rule based engine that triggers/fires the event when the arrived sensor data readings meet a pre-defined rule/condition (i.e. equal to, more than, less than, etc.). Some people refer to rule based engine as IFTT (If This Then That) or if this data reading arrived then do that event/action (send Email, send SMS, call external web service, etc.). The same module is responsible for data enrichment as well as data orchestration data orchestration is the distribution of data to other external systems in a certain flow.

D- Administration and access rights management module: this module is required to manage both end users' and developers' accounts and their access rights towards the various projects, applications and IoT devices. this module is also responsible of managing devices credentials, encryption keys and data acquisition protocol details. The same module allows the administrator to manage several settings of the IoT AEP including the AEP built-in MQTT broker, content policies, etc.

Finally, a true IoT AEP empowers IoT solutions with a neat solution architecture and design as the true IoT AEP should also provide the developers with reliability, scalability and performance. Have you ever thought ?? of your IoT platform reliability? and what secretes your platform provider doesn't want you to know ?

It is worth mentioning that lots of the big ICT companies have failed in IoT AEP business, regardless of being well known brands and having annual big marketing budget that paid a lot to market analysts to market their platforms, have failed in this category and dropped its business line in this IoT AEP category for various reasons that we have discussed in multiple LinkedIn posts. Those who failed in IoT AEP and dropped it include Microsoft failure in IoT , 思科 Kinetic, 谷歌 Cloud IoT core, the fake IBM BlueMix IoT and the fake IBM Watson IoT :-)

Discussion on failures of some brands are available in the links below

• Poor Market Analysts create IoT platforms reports according to brand power.
• Microsoft is additional one more giant (elephant ??) that fails on IoT with Azure IoT central.
• Microsoft is trying to fix and correct their Azure IoT failure in IoT business
• IBM killed its fake Watson IoT platform effective December 1st, 2023.
• Google announced a planned end of life for Google IoT cloud core in August 2023.
• Honeywell failure stories in IoT.
• Gartner is confused/mixing expectations from "IoT Device (Thing) Management" technology and "IoT platforms" technology.

5-2 Open-source IoT AEP family:

An open-source family that offers free community version and an enterprise licensed version. This category focus more on providing a mix of libraries for free use by the developers to use to develop their own applications. The best in this category is ThingsBoard and it is developed by a company in Ukraine. The most complex in this category is FIWARE as an abbreviation for Future Internet Ware and it is developed as financed R&D contributions from H2020 by various European entities including companies, research centers and universities. All these entities are financed to add enhancements or features as much as they like to add to FiWare and that made FiWare a complex mix of too many libraries for whatever task or purpose you might need to develop, hence, FiWare itself became an undefined software and considered to be a joke in the community of IoT Platform makers :-) as the future undefined software.

5-3 IoT AEPs by SCADA providers:

This IoT AEP family is owned by SCADA (Supervisory Control and Data Acquisition) solution providers where they all focus on IIoT (Industrial IoT ) that will replace SCADA systems. Almost all of the successful and the big SCADA providers decided to acquire an IoT platform before the IoT kills their scada business. PTC acquired ThingWorx, Schneider Electric Industrial Automation acquired WonderWare, and 西门子 acquired MindSphere. Software AG acquired Cumulocity for Industrial IoT as well. These IIoT platform providers have invested in these acquisitions to keep their place in industrial automation and utility grid management but definitely not suitable for smart city use cases (i.e. street lighting, parking management, waste management, transportation management, smart utility metering, etc.). On the contrary 通用电气 (General Electric) wanted to build their own IIoT platform and had a well known failure story because they decided to build their own Predix IoT Platform by their own with no knowledge or experience in IoT and assuming it is similar to SCADA. GE ended up with huge cost under the dying business line of GE Digital.

5-4 IoT AEPs by database engine makers:

This IoT AEP family is owned by Database engine makers. Some of the database engine makers like SAP acquired software that can collect data from IoT devices, usually through MQTT, and store it inside the database engine and that allows the database engine makers and their community of partners to use their existing development tools to develop applications based on the data collected from the devices and stored in the database as they used to do in any database application.

An example of this category is SAP that acquired PlatOne and branded it as SAP Leonardo. Later SAP phased out/killed Leonardo silently without any announcement and just removed it from their web site and started relying on 3rd party platform provider companies that are mainly focusing on IIoT (industrial IoT) use cases through cumulocity platform. We are happy to see SAP passing over its failure in IoT with Leonardo and trying to catch up in IIoT (Industrial IoT) through partnership with Cumulocity IIoT.

5-5 IoT AEP family that originated from Messaging bus:

Providers in this family tried to implement the messaging bus technology, that is originally used for messaging (to enable communication) between various systems inside complex IT environments (i.e. the IT environment in mobile telecom operators and banking institutions) as a messaging bus, for devices to communicate to each other and communicate to surrounding servers. Carriots is a Spanish company that started this category and later acquired by Altair and then Altair replaced it by SmartWorks. Carriots represent the failing IoT messaging bus category.

6- Analytics platforms:

These are platforms used to analyze all the big data collected over many years from various IoT applications running in a city or a large corporate/enterprise to figure our new relation between the collected data sets. Hadoop is the most well known open source big data analytics platforms and the rest are licensed commercial software.

7- Device & gateway remote management platforms:

These DMP (Device Management Platforms) are platforms used to remotely configure the IoT devices and IoT Gateways to send remote software updates (device firmware) to these IoT devices and IoT gateways. Compared to the mobile value chain, this is similar to the Play store functionality in Android mobile phones but for IoT devices.

Some of the IoT devices are doing multiple functions each function is separate in a software application and hence these devices are running RTOS (Realtime Operating System) comparable to the Android OS on your phone and in this case the device RTOS will be running a group of applications on top of it like the applications installed on top of your Android phone or applications installed on our iOS if you are Apple phone user. Some IoT devices are doing very basic functionality and hence it is using a firmware that directly interacts with the device HW without relying on an RTOS.

Whether the IoT device is running applications on top of RTOS or a simple firmware, there will always be an need to update the device firmware to add new features or fix/correct a bug in that device software. Because IoT devices are running remotely in the filed in customer sites and geographically distributed, you can not go and visit every device and attach it to a computer to upgrade the device software. You need a way to remotely manage and update all these IoT devices from a central location and the cloud based software doing this is called DMP (Device Management Platform).

The process of remotely updating the device firmware is usually referred to as FOTA (Firmware Over The Air) or simple OTA (Over The Air) referring to the process of sending (pushing) information to the device to start downloading a new firmware.

Kaa is an open source device management platform while the rest are licensed commercial software. Kaa provides a list of SW libraries (i.e. C files) to use inside your device firmware to receive updates. These Kaa libraries differ from RTOS to the other depending on which RTOS is used in device.

Balena is more innovative in doing the device management through a container like technology to remotely update the image on the device and this implies the device should have an RTOS with the container management capabilities to receive the new updated device SW image whenever there is a new device SW image released.

Some IoT AEPs like MasterOfThings include an application to do remote device management utilizing MQTT without the need to a dedicated remote device management platform.

8- End-to-End system integration companies

These are companies that have competent staff of resources who have skills and knowledge in every building block/component of the IoT Value chain. Those are the companies contracted by the customers to develop and provide complete end to end solution utilizing a mix of the IoT value chain components. The most important skill that must exist in these companies is the solution architect role then comes other competent resources including the application developers and system integration engineers (or service engineers) and the project managers.

Details of these roles and skills required for each role are explained in the free online career talk 10 minutes video available on MaharaTech learning portal in the link below.

https://maharatech.gov.eg/course/view.php?id=1325