Low-code adoption should be measured by business outcomes rather than the number of applications created’

Dr Utpal Chakraborty, AI & Quantum Scientist; Founder & Chief Scientist, GAHNA AI Lab (ExorionAI).

How are low-code automation platforms changing the way enterprises approach Industry 4.0, particularly by enabling engineers, plant teams and business users to participate directly in developing automation applications?

Industry 4.0 is fundamentally about connecting machines, data, intelligence and human decision-making. However, one of the biggest barriers to its adoption has been the dependence on specialised software developers and automation engineers to translate operational requirements into digital solutions.

Low-code platforms are changing this equation by democratising application development.

A production engineer who understands machine behaviour, a maintenance supervisor who knows recurring equipment failures, or a quality engineer who identifies process deviations can now participate directly in building digital workflows using visual interfaces, preconfigured components and reusable connectors.

Consider a manufacturing plant where equipment downtime is recorded manually. Traditionally, developing an automated monitoring and escalation application might involve multiple teams, lengthy specifications and considerable custom programming. With a low-code platform, the plant team can help create dashboards, define alert thresholds, configure maintenance workflows and automate notifications with significantly less coding.

This represents an important shift from IT-driven automation to business-led innovation supported by IT and OT governance.

However, low-code should not be confused with unrestricted access to industrial control systems. Safety-critical PLC logic, emergency shutdown systems and deterministic real-time control must remain under qualified engineering supervision.

I believe the real significance of low-code lies in bringing application development closer to the people who understand the operational problems best.

What impact can low-code platforms have on application development speed, deployment and time-to-value?

The traditional enterprise application lifecycle involves requirement gathering, development, integration, testing and deployment. In industrial environments, these stages can become particularly time-consuming because multiple technologies, vendors and operational dependencies are involved.

Low-code platforms can substantially compress the application-development portion of this cycle.

For example, a manufacturing company identifying excessive energy consumption across production lines could create an application that consolidates energy-meter data, visualises consumption patterns, flags anomalies and generates recommendations without developing every software component from scratch.

A focused proof of concept that might otherwise require several development iterations can potentially be demonstrated within days or weeks, depending on integration complexity.

More importantly, low-code supports iterative innovation. Plant teams can rapidly test an idea, collect feedback, refine workflows and expand successful applications to additional production lines or facilities.

This becomes even more powerful when combined with Artificial Intelligence.

Imagine a low-code application integrated with a predictive-maintenance model. Equipment telemetry is continuously evaluated, potential failure conditions are identified, and maintenance requests are generated through an automated workflow.

The enterprise moves from reactive maintenance towards predictive operations.

Nevertheless, development speed should not be confused with production readiness. Security testing, operational validation, integration certification and change-management procedures remain essential.

The real value of low-code is not merely writing applications faster; it is reducing the time between identifying an industrial problem and delivering measurable operational improvement.

How can low-code platforms bridge specialised automation and IT skills gaps while empowering the existing workforce?

The shortage of professionals who simultaneously understand operational technology, industrial processes, software engineering, data analytics and AI remains a major challenge for Industry 4.0 adoption.

Low-code platforms can help address this challenge by enabling a broader category of professionals to contribute to digital transformation.

A mechanical engineer need not become a full-stack programmer to develop a machine-performance dashboard. Similarly, a quality-control specialist should not require extensive software training to configure a digital inspection workflow.

This creates opportunities for what I call domain-led citizen development, where employees with deep operational knowledge build applications within an approved enterprise technology framework.

However, successful adoption requires structured workforce development.

Employees should understand data fundamentals, workflow modelling, APIs, application testing, cybersecurity and the limitations of automated decision-making.

Enterprises can establish internal low-code competency programmes, supported by experienced IT and OT professionals who review architecture, security and maintainability.

AI-assisted low-code development can further accelerate this transformation. Generative AI can help users describe requirements in natural language, suggest workflows, generate interfaces and explain application logic.

But AI-generated configurations must still be validated before deployment.

My perspective is that low-code should not replace skilled developers or automation specialists. Instead, it should enable those specialists to focus on sophisticated engineering challenges while empowering operational teams to solve routine digital problems independently.

How can enterprises ensure seamless integration with PLCs, SCADA, MES, ERP, IoT and other Industry 4.0 systems?

Interoperability is arguably the most critical architectural consideration for enterprise-scale low-code adoption.

Most industrial organisations already operate heterogeneous technology environments involving PLCs, SCADA, Distributed Control Systems, Manufacturing Execution Systems, ERP applications, industrial IoT gateways and enterprise databases.

Introducing low-code applications without a clear integration strategy risks creating additional technology silos.

The solution is an API-first, standards-based integration architecture.

Industrial communication standards such as OPC UA, along with protocols such as MQTT and established enterprise integration interfaces, can provide connectivity between equipment, operational platforms and business applications.

ISA-95 concepts can help maintain consistent information models between manufacturing operations and enterprise systems.

Low-code applications should ideally interact through governed integration services, APIs or industrial edge gateways rather than directly accessing sensitive controllers.

For instance, a production optimisation application could consume equipment performance data through an OPC UA-enabled industrial gateway, correlate it with MES production schedules and ERP inventory information, and recommend adjustments to operational planning.

Such an architecture enables information exchange without unnecessarily exposing underlying control infrastructure.

Enterprises should also standardise data definitions, asset identifiers, timestamps, units of measurement and application interfaces.

For industrial organisations, the objective should be to establish one connected digital ecosystem rather than hundreds of disconnected applications.

What governance, cybersecurity, data-management and quality controls are essential when automation development becomes distributed?

Democratising development introduces significant opportunities, but it also expands the governance challenge.

An incorrectly configured workflow in an office environment might delay an approval. A similar mistake affecting industrial operations could disrupt production, compromise equipment availability or introduce safety risks.

Consequently, low-code governance must be designed into the platform from the beginning.

Enterprises should implement role-based access control, strong authentication, application ownership, approved connectors, version control, audit trails and mandatory testing before deployment.

Every application should have a clearly defined purpose, accountable owner, data-access permissions and lifecycle-management process.

From a cybersecurity perspective, alignment with the ISA/IEC 62443 framework is particularly important for industrial automation and control environments.

IT and OT networks should remain appropriately segmented, sensitive integrations should use secure gateways, and applications must operate according to least-privilege principles.

AI-enabled workflows introduce additional considerations, including model reliability, data leakage, prompt injection, explainability and human oversight.

For example, an AI-driven application may recommend changing maintenance schedules based on predicted equipment failures. However, recommendations affecting safety, production-critical equipment or regulatory compliance should require appropriate engineering approval.

I strongly advocate the principle of controlled democratisation: decentralised innovation with centralised security, architecture and governance standards.

Low-code should expand an enterprise's innovation capacity without expanding its cyberattack surface uncontrollably.

How should enterprises measure success, and what does a successful enterprise-wide low-code roadmap look like?

Low-code adoption should be measured by business outcomes rather than the number of applications created.

Relevant metrics include development lead time, deployment frequency, maintenance effort, application support costs, employee productivity, production downtime, Overall Equipment Effectiveness (OEE), Mean Time to Repair (MTTR), energy consumption and return on investment.

For example, if a low-code-enabled maintenance workflow reduces fault-response time and prevents avoidable production interruptions, the resulting operational savings provide a stronger measure of success than the development cost alone.

I recommend a phased enterprise roadmap.

• Phase 1 – Identify high-value opportunities: Select practical use cases such as digital inspections, predictive-maintenance notifications, energy monitoring, quality reporting or maintenance approvals.

• Phase 2 – Establish the technology foundation: Define approved platforms, integration standards, cybersecurity controls, data architecture and application governance.

• Phase 3 – Execute controlled pilots: Develop a limited number of applications with measurable performance baselines and clearly defined operational ownership.

• Phase 4 – Build workforce capabilities: Train engineers and business users, establish an internal Centre of Excellence and introduce reusable application components.

• Phase 5 – Scale and optimise: Expand successful applications across plants, integrate AI-driven insights and continuously evaluate operational and financial outcomes.

Looking ahead, I see low-code evolving into an important foundation for AI-powered Decision Intelligence and intelligent industrial operations.

Industrial systems will increasingly move beyond simply monitoring equipment towards interpreting operational context, anticipating disruptions, simulating possible responses and recommending corrective actions.

Low-code platforms can provide the application and orchestration layer connecting these intelligent capabilities with human operators and enterprise systems.

However, autonomous execution must remain bounded by engineering safety constraints and enterprise-approved controls.

Ultimately, the future of Industry 4.0 will not depend exclusively on how many software engineers an organisation employs. It will depend on how effectively the organisation enables its engineers, operators and business professionals to translate industrial knowledge into intelligent digital solutions.

The next industrial revolution will be driven not merely by connected machines, but by empowered people, intelligent decisions and democratised innovation.

(The views expressed in interviews are personal, not necessarily of the organisations represented.)

Dr Utpal Chakraborty (PhD) is an AI & Quantum Scientist and the Founder & Chief Scientist of GAHNA AI Lab (ExorionAI), working on sovereign AI, cybersecurity, Decision Intelligence and industry-specific intelligent systems. He previously served as Head of Artificial Intelligence at YES Bank and held senior technology and architecture roles associated with IBM, L&T and Capgemini.

A recognised AI thought leader, researcher and international speaker, Dr Chakraborty actively contributes to discussions on responsible AI, Agentic AI, industrial digital transformation, cybersecurity, quantum technologies and the future of intelligent enterprises.