Life is driven by rhythms. Everything in this world seems to be cyclic/periodic in nature. In doing so each cycle comes with its own past learnings and improvements and vastly changes us and our lives for better. All of us with life-sciences background know this evolutionary process.
Computing industry also seems to be following this evolutionary cyclic pattern as shown in the graphic above of going from central to distributed to central and then again towards distributed computing. Each iteration averaging around 2-3 decades. And every iteration coming strong, disruptive and much better than the previous iteration. All the learnings of earlier phases of the computing paradigm cycle helps create the new and improved computing paradigm.
We all know that personal computer and then mobile phones were such a huge paradigm shift from mainframe computing that it had helped bring computing in the hands of common man as a commodity. Cloud computing has even further helped penetrate computing to the masses to make it like a utility (internet: similar to how we consume water or electricity). However, cloud wouldn’t have been so successful without the past learnings from mainframe and personal computing experiences and also without embracing open source software especially the Linux operating system and its FOSS ecosystem. Now almost everything cloud-native is being developed and runs on Linux or its derivatives. Furthermore, the ability to add computing resources and storage on-the-fly like a commodity without getting to the hassle of background infrastructure programming, management and maintenance was such a huge advantage for general purpose computing in the consumer space and peripheral business computing. The cloud offerings for domain agnostic applications in supply-chain, logistics, GPS, CRM, word-processing, email etc have worked wonders for both end-consumers and businesses alike.
However beyond the peripheral non-mission critical applications, cloud has been difficult to penetrate into domain specific mission critical applications and infrastructure needs. This is especially true in process manufacturing space where core manufacturing processes are still executed with traditional on-premise computing gear (with local data centers) near the process machinery on the plant floor. This is more true for the regulated industry like pharmaceuticals and biotech. The low adoption of central cloud based offerings on the manufacturing floor stems from:
- Need of very low latency
- operational autonomy
- separate IT and OT network and connectivity protocols
- near real-time decision making on manufacturing
- extremely high volume of time-series data generated by these industrial machines making it uneconomical to send everything to public cloud
- need for local data storage, accessibility and analytics
- poor internet, wireless and cellular network connectivity makes it harder to have all computing and storage on the public cloud
- regulatory and other business needs to keep information on-premise
Efficiencies gained from cloud applications in the generic consumer and peripheral applications can easily be translated into mission critical domain specific process manufacturing if we can bring cloud-native technologies on-premise. If industrial production staff is able to get same computing experiences as we get on the public cloud which provides
- easy scalability
- securely interconnect IT and OT world
- is highly available and fault tolerant
- no-code plug’n’play with domain specific applications
- act as local data storage and process historian
- and satisfy the local/distributed computing needs
the industry will tremendously benefit and truly achieve the Industry 4.0 goals of making data driven highly interconnected plant floor. With the evolution of technologies like message brokers, document databases, observability tools, containerization etc, the manufacturing industry has direct applications of these cloud-native technologies to interconnect IT and OT world for real-time data driven decision making right next to the process machinery.
We at Factorian are busy building cloud-native edge products (tightly coupled with autoscaling computing hardware) that will be trying to solve these needs primarily focused on pharmaceutical, biotech and specialty chemicals industry within the regulatory framework.
If you wish to be an early adopter do get in touch with us to know more.