THANK YOU FOR SUBSCRIBING
Building an IoT Platform Supporting a Digital Business Model
By Kalman Tiboldi, Chief Business Innovation Officer, TVH
IoT is transforming a world of things into a world of data. More products are natively built with embedded capability to collect, select, analyze and transfer data becoming in this “smart” way. In the domains where TVH Group offers maintenance and repair related services as material handling, industrial and agricultural equipment the machines are gradually evolving toward smart mobile machines, including AGVs, with cognitive computing systems on the board that can make decisions and solve problems with limited human intervention. IoT enabled data driven predictive maintenance is becoming relevant in all these industries as it can drive efficiency by providing higher levels of safety and quality at a fraction of the current costs. Thanks to Advanced data analytics and IoT devices, predicting potential failures is gradually becoming a real capability.
In order to boost the performance of mobile machines used by TVH customers, we have decided to build a new company called GEMone, running a digital business supported by a technology platform. GEMone is using a subscription-based business model, bundling physical and digital products and offering services in four key domains: Track and Trace, Security and Safety, Operational efficiency and Service and Maintenance.
An IoT and data driven digital business requires much more than technology for example: organization, culture, leadership, talent and skills, and a new business model, but the IoT technology platform supporting the business is one of the major components. There are multiple often overlapping definitions of IoT platforms. I have simplified and applied it to our business case, looking to it as a set of integrated hardware, software and data capabilities w hich allow the collection, transfer, storage and analytics of data generated by our connected machines.
One of the biggest decisions facing many companies starting IoT projects is whether to build IoT platform or to adopt an existing one
The data and intelligence created in this way are used by various digital products and services. Major capabilities I am considering for the IoT platform are:
• Edge devices management
• Data management
• Data analytics and models
• Business application enablement
• API integration and ecosystem support
• Security and compliance management
Close to these capabilities, key features like availability, scalability, ease of implementation and use, low running and maintenance costs are important as well.
One of the biggest decisions facing many companies starting IoT projects is whether to build IoT platform or to adopt an existing one. In our case, after evaluating different existing platforms, we have decided to go for the build option. We have designed a multi-layered reference architecture and started to build the platform using some well-known software and data architecture principles:
• PaaS solution built on the top of Google GCP
• Managed, native data components like Pub/Sub, DataFlow, DataProc, BigTable, BigQuery, Cloud ML
• Server-less computing
• Microservice-driven software architecture
• Real-time event driven service architecture
• Multi-protocol and device based data ingestion
When it comes to building an IoT platform, there are many challenges and technical hurdles. The good news is, even if your company has diverse requirements you don’t have to build an IoT platform from scratch. There are many ready-to-use building blocks delivered by cloud vendors like Amazon, Google, and Microsoft. You can choose and stitch them together. However, if you choose for natively integrated and managed components, your cloud portability becomes a problem. For functional components you are deciding to build yourself, you can gradually get business benefit from the platform using Minimal Viable Product (MVP) development. By adopting a product oriented agile organization structure and agile software development methodology, by working in short iterations with strong focus on customer experience and business value delivery, by presenting MVPs to our test customers, we managed to create business benefits and get our services validated in an early stage. Going through this process and working closely with external companies and experts, including Google, we managed to build internal expertise in diverse domains required by our digital business like: R&D, architecture, UX-design, software development, data management and analytics including the running and maintaining of a server-less environment.
By building our platform we have been facing some major technical challenges like dealing with high volume and frequency of data, building the right edge analytics capability, combining batch and streaming analytics, and security and compliance. However at the same time, we managed to deliver the expected business value by connecting more than 25000 machines to it. The platform became in this way a major accelerator for innovation and a real ecosystem for collaboration with customers and partners.