It was four years ago that I decided to open source my time-series database project, TDengine, and I’ve never regretted it. At the time, my friends and peers told me that making my most valuable asset — the source code to my product — available to the public for free was a recipe for disaster. Why would anyone purchase a product they could compile themselves at no cost?
However, I was a developer myself before I became an entrepreneur, and I knew that for TDengine to succeed, it needed to make developers succeed. We needed people to use, contribute to and advocate for the product — whether in their companies, educational institutions or wherever time-series data was being used. Only by giving back to the developer community (in the form of open source) could we expect to win the market.
Four years later, TDengine has been installed over 400,000 times and has more than 22,000 stars on GitHub. The product and company are stronger than ever thanks to the support we’ve received from developers. Over the past year, however, I have realized that my original plan to empower developers did not go far enough.
Having learned more about the industrial data scenarios where my product was being used and having spoken with many experts in various fields, I’ve realized that industrial data must be democratized with the power of open source.
Democratizing Industrial Data
While open source has become the mainstream in the high-tech world, the opposite is true in traditional industries. Industrial data platforms are still living in the past, depending on closed systems and vendor lock-in to continue milking profits from decades-old software.
For existing enterprises, replacing these outdated but necessary systems is too costly to consider in terms of both time and money. For startups, the high barriers to entry and unfriendly pricing plans of these “industry-standard” platforms have likely prevented numerous companies from being successful.
With the advent of Industry 4.0, the IIoT and now AI, everyone in traditional industries like energy and manufacturing is looking for a path forward. However, the very companies that produce the current mainstream industrial data platforms are blocking that path. These companies depend on their closed nature to keep their customers; if you can’t get your data out, you can never move to a competing product.
Integration, standardization and centralization are necessary for modern data management and analytics, but these are not priorities for industrial data platform vendors — precisely because these concepts and technologies, which are necessary to move the industry forward, help existing customers escape the iron grip of current vendors.
In short, legacy systems are roadblocks to digital transformation, and the only way around these roadblocks is through openness. The open-source model can make industrial data accessible to everyone. A rich ecosystem of open systems can decouple traditional industries from the outdated historians that have held them back for so long and introduce collaboration and integration on an unprecedented scale.
How You Can Prepare
If the future of industry is indeed open, industry leaders may wonder how they can prepare themselves and their organizations to benefit most from it. Open-source systems and cloud computing may revolutionize infrastructure, but they will not make existing teams and processes obsolete — provided those teams can adapt themselves. Here are a few tips to help you obtain the most benefit.
- Insist on standard, open interfaces and protocols. Being locked into systems makes it so difficult to migrate to or integrate with other products that it’s easier to simply maintain the status quo — even when alternative solutions may better serve you. Going forward, ensure that new data infrastructure purchases run on standard interfaces and protocols such as MQTT so that you can more readily switch out components that aren’t the best fit for your company.
- Choose products that can be run either on-premises or in the cloud. The future of industrial data might be in the cloud, but are you ready to make that jump all at once? While the SaaS model may seem like a good way to reduce your expenses, some enterprises are actually moving back to physical data centers where they can keep costs under control. For the time being, the best choice is to select systems that offer cloud and locally deployed options so you can decide on your own terms when and what components to migrate.
- Look for comprehensive solutions instead of systems with too many moving parts. It’s all too common that open-source systems involve a complex combination of multiple products, requiring significant technical expertise that your team simply doesn’t have. When you add up the time spent learning how to use, integrate, and debug these various components, it’s not uncommon to realize that the overpriced monoliths of the past may have actually been more economical. Instead, focus on products that provide the services you need as a consolidated package to reduce the complexity of your systems and lower the learning curve.
- Don’t be afraid to try new things. With open-source and cloud technology, you benefit from a great variety of options. That latest and greatest AI analysis tool might be your key to operational efficiency — or it might be completely useless in your business scenario. In the SaaS world, there is a wealth of alternatives that you can try out at little risk of cost or time, so make sure to consider all options.
Open source and the cloud are prepared to disrupt the field of industrial data. Let’s get ready together. I’m excited to see the concepts of IIoT and Industry 4.0 finally brought to fruition and benefiting the entire industrial sector.
This article was originally published in Forbes Technology Council.
