Finding USPs in the IoT ecosystem: Part 2

Arpit Khosla of IoT Australia Consulting

Via this series of articles, our objective is not only to chart out the IoT Value chain but also the market/economic structures and changing dynamics around them. This write-up is aimed at addressing the business decision makers who are attempting to figure out a solution strategy & USP in each component of Value chain. In our last article we focused on IoT sensors, devices, gateways & communication part of Value chain. In this article we, Arpit Khosla and Praneet Thakur of IoT Australia Consulting Services shall dive into IoT Platforms.

Considering the multitude of ways in which the term “IoT Platform “can be interpreted, we take this opportunity to first give our view on what all does an IoT platform encompass. At a high level, it encompasses horizontal & vertical capabilities. Some of the capabilities that can be flagged as horizontal are Connectivity Management, Device Management, Data Ingestion & Storage, Application Enablement Platform & Analytics enablement Platforms. Sometimes Billing platforms providing End to End Billing are also a part of the Horizontal Platforms Family. The Applications themselves are more vertical or scenario specific and hence can be categorised as being a vertical capability. Let us look at each of those capabilities in from strategy & USP lens.

As far as Connectivity Management is concerned, though one may need to choose the right technology based on coverage range, security, mobility, throughout, Scalability, power profiles & latency requirements, but solution itself is typically taken care of by network providing operator. From solution strategy perspective ,there are some key decisions like licensed vs unlicensed SIM, SIM vs eSIM, & split billing vs end to end Billing but they warrant another Article, so we will not go in their details in this one.

Similarly from device management solution strategy perspective the technical decesion maker needs to address the requirements of Device lifecycle management, command and control, monitoring, auditing, firmware updates at. In this context the decision maker needs to consider not only standard protocols like LWM2M & OMADM but also context specific protocols which could be OPC-UA, Profinet, Modbus, and Hart etc. However from USP perspective, there is not much that an Innovator can draw from either connectivity management or device management functionality.

The data ingestion & storage platforms are no strangers to challenges of heterogeneity, evolving standards and requirement to handle scalability with the right security posture. From a solution strategy perspective, choosing the right protocol amongst https, MQTT, CoAP, AMQP etc. is one key decision to make. Typically most platforms offer support for above protocols but it is the commercials and ease of integration that drives the decision here. For instance some platforms provide set of libraries to enable communication which warrant time & effort investment for development, on the other hand others offer low code / low touch configuration based setup.

From Solution availability perspective we see three approaches. Firstly data ingestion brokers and storage platforms are part of suite offered by End to End Horizontal Cloud platform providers such as AWS & Azure. These attempt to simplify the complexity of Infrastructure setup & management, scalability, high availability etc. An alternate solution is to buy fully productised platforms such as C3IoT, Software AG, PTC. By productised we mean typical production requirements around administration, operation, auditing, integration and user interface are pre baked.

The typical third option is to build using open source solutions like Rabbitmq, Eclipse Mosquito etc. Overall most PaaS & SaaS players have similar offerings in this space and typical considerations for making choice are scale, commercial, operational and ongoing maintenance & support. We do not see this part of the solution as a strong candidate for building any core differentiator or USP, additionally with plethora of solutions available we expect this part of value chain to remain competitive and hence armed with little chance of PaaS or SaaS abusing market power. Conclusively we could mark this as a competitive and commoditised space, with PaaS or SaaS being the preferred mode of development.

Praneet Thakur

Application enablement platforms and Analytics platforms can as well be categorised in the horizontal platforms space. These platforms enable your solution to deal with incoming data in real-time or batch / scheduled manner. This is different from data ingestion platform as here not all platforms have similar capabilities. As a technical decision maker one usually evaluates on the capability of performing Analytics and Machine Learning, visualisation application development & hosting capabilities. Extensibility via Integrations and API calls are another dimension of consideration. These integration can be with environments which are monolithic or micro services based, Cloud or On-premise based.

Some of these solutions also foray into Edge space especially for edge analytics. Once the decision maker knows the requirements from above dimensions perspective, yet again, he can look at either of the three choices PaaS offerings from Azure & AWS or Open source based build. Considering that the need for edge or fog computing is growing fast and considered as the most growing area, it is almost mandatory for the solution designer to have a clear Edge strategy as well. In Edge space one could look at solutions of the likes of Intel Movidius based chipsets & edge appliances by Dell from hardware perspective and Azure IoT Edge, AWS IoT Greengrass, Software AG edge appliance from software perspective. In analytics and application space while following the third approach of build one has to stitch technologies like Kaa, HDFS, Kafka, Nifi, Mongo DB, Nginx etc. together to enable a robust IoT Solution.

Overall from solution strategy perspective we believe one needs to be aware of the fact that SaaS players in analytics space are expected to become differentiated with optimised learning algorithms along the way. This creates the possibility of SaaS players abusing the market power latter in the day. The Business decisions maker also needs to be cognizant of the time and investment required in build approach, which might not match up to the differentiation developed in this space. Hence we believe the preferred approach here is PaaS for Application enablement & Analytics engine functionalities.

Next in line are the Vertical specific applications. The Solution builders will typically call various functionalities from the above catalogue of IoT Platforms to build the end to Application. For instance –the application could be developed in an environment provided by Application enablement platform by collecting data from data ingestion platform, and thereon developing visualisation on top of the trends & insights mined by Analytics platforms. It is when the insights generated by this end to end context based stitching of ecosystem, solves a problem, then one knows that value has been created and thereby arming the solution with a strong USP.

This is the area which needs examination with relatively much larger business lens. The decision makers need to evaluate if the application is valuable, rare, imperfectly imitable & non substitutable. From market structure perspective this is an area where one can define his own market and enjoy monopoly, hence a build from scratch possibly using open source environments is a preferred approach. Additionally from business perspective, we expect high stickiness of customers in this space, hence time is of essence here, and the sooner one can go to market with the chosen solution, the higher the chances of leveraging the USP to scale.

Overall from a USP development perspective, it is the Vertical specific Applications with underlying customised Analytics that seem to be most suitable candidate. From an IoT Platform solution strategy perspective though we have given recommendations of PaaS or SaaS for Data ingestion, PaaS for Application and Analytics and Open Source based build for Vertical specific applications perspective, but we appreciate the immense importance of the context that drives all decisions. If the context warrants rapid time to market with support of a dedicated R&D : Saas might become an optimal decision. Hence all above factors can be used as a general guideline, but each use case will needs its own Jury and its own decision.

The authors of this blog are Arpit Khosla, founder of IoT Australia Consulting Services and Praneet Thakur, advisor of IoT Australia Consulting Services 

Comment on this article below or via Twitter: @IoTNow_OR @jcIoTnow

Recent Articles

AspenTech expands application of industrial AI

Posted on: May 12, 2021

Aspen Technology, Inc., a global provider in asset optimisation software, announced that it has extended Industrial AI across its solutions to drive higher levels of profitability and sustainability in customer’ operations.

Read more

ZEDEDA signs strategic OEM supplier agreement with Agora

Posted on: May 11, 2021

ZEDEDA, the provider in orchestration for the distributed edge, announced an integration with Agora that provides customers in the oil and gas and renewable energy industries with full lifecycle management capabilities for their edge deployments as part of Agora’s overall IoT solution portfolio.

Read more