Bringing datasets together creates a much richer picture of the production environment

Sameer Gandhi, Managing Director, OMRON Automation India.

How are advances in industrial sensors and machine vision changing the way smart factories capture, interpret and use real-time information from the physical production environment?

The role of sensing in manufacturing is undergoing a significant transformation. Traditionally, sensors and vision systems were primarily used to detect a specific condition or object, measure a parameter or identify a defect. Today, they are becoming increasingly important sources of contextual information about what is happening across the production process.

Modern sensors can capture increasingly precise information at higher speeds, while machine vision has evolved to inspect, measure, identify and interpret complex visual conditions. When this information is available in real time, manufacturers can move from simply knowing what happened to understanding what is happening and, increasingly, why it is happening.

At OMRON, this evolution is closely aligned with our core technology concept of “Sensing & Control + Think.” Sensing is about acquiring the right information from the real world; Control enables the production system to respond; and “Think” brings intelligence into this cycle so that data can be interpreted and converted into meaningful action.

The important shift is therefore not simply towards collecting more data. It is towards collecting the right data, at the right time, in the right context and the using it in an efficient way.

For manufacturers, this creates the foundation for more responsive and adaptive production. A vision system, for example, can identify a quality deviation, while process sensors and machine data can provide additional context. The combination enables faster intervention, better process control and, ultimately, more consistent quality.

As Indian manufacturing becomes increasingly sophisticated and globally competitive, this ability to convert physical phenomena on the shop floor into actionable digital information will become a critical competitive advantage.

What is the significance of combining data from sensors, machine vision, PLCs, robots and other industrial systems, and how does data fusion improve the accuracy and reliability of manufacturing decisions?

The real value of industrial data emerges when different sources of information can be viewed together rather than in isolation.

Consider a simple example. A vision system may identify a dimensional or assembly defect. On its own, that tells us what happened. If we combine that information with PLC data, motion information from a robot, sensor values, process parameters and production history, we can begin to understand why it happened.

This is where data fusion becomes important. Different technologies provide different perspectives of the same manufacturing process. Sensors measure physical conditions, vision provides visual information, controllers provide process and machine information, while robots and motion systems provide information about movement and execution.

Bringing these datasets together creates a much richer picture of the production environment. It can improve anomaly detection, enable more accurate root-cause analysis and support faster decision-making.

This also represents an important evolution in factory architecture. Manufacturing systems are increasingly moving from predominantly control-centric architectures towards more data-centric approaches, driven by developments in IoT, AI and connected technologies. OMRON's recent technical work specifically addresses this evolution and the need for flexible, data-oriented factory automation architectures.

However, data fusion should not mean creating a massive data repository without a clear purpose. The objective should always be to connect data around a specific manufacturing needquality, productivity, maintenance, traceability or energy efficiency and convert that information into an actionable decision.

Ultimately, better-connected data leads to better-informed decisions, provided the data is accurate, contextual and available at the point where the decision needs to be made.

How is machine vision evolving from a conventional quality-inspection tool into a broader source of process intelligence – for example, in assembly verification, process monitoring, traceability and predictive quality?

Machine vision has evolved considerably beyond the traditional concept of simply asking, “Is this product OK or not?”

Today, vision can provide information throughout the manufacturing process. It can verify whether the correct component has been assembled, confirm orientation and positioning, read codes for traceability, monitor process conditions and identify deviations that may otherwise be difficult for the human eye to detect.

This creates an important opportunity: vision can become a source of process intelligence, rather than just a final inspection mechanism.

For example, identifying a defect at the end of a production line is valuable, but identifying a visual pattern that indicates a process is beginning to move out of specification is even more valuable. It gives manufacturers an opportunity to intervene before the problem results in significant scrap, rework or downtime.

The combination of machine vision with AI and other manufacturing data can further enhance this capability. OMRON's technology development includes AI-based visual inspection and applications aimed at detecting potential defects and supporting quality management.

We therefore see the role of vision expanding from inspection to understanding.

For industries such as electronics, automotive, batteries and other high-precision manufacturing segments, this is particularly important because quality requirements are becoming more stringent.

The future of machine vision will not be defined simply by higher camera resolution or faster inspection. Its real value will come from how effectively visual information can be connected with process information and used to improve the manufacturing process itself.

How can the combination of sensor data, visual information and historical production data help manufacturers detect anomalies early and move from reactive maintenance towards predictive and even prescriptive maintenance?

Maintenance is another area where the combination of different types of data can fundamentally change the way factories operate. Traditional maintenance is often either reactive repairing equipment after a failure or preventive, where maintenance is performed according to a predefined schedule. Both approaches have limitations. Reactive maintenance can result in unexpected downtime, while preventive maintenance may mean replacing components before they have actually reached the end of their useful life.

By combining real-time sensor information, machine conditions, visual information and historical production data, manufacturers can begin identifying patterns that precede an abnormal condition.

For example, a change in vibration, temperature, current, cycle time or visual characteristics may appear insignificant when considered individually. When these signals are analysed together and compared with historical patterns, they can provide an early indication that equipment or a process is beginning to deviate from its normal condition.

This is where AI and edge computing can play an important role. OMRON has developed AI-equipped machine controllers that enable machine learning and anomaly detection close to the production process, allowing data to be analysed at high speed and used for real-time decisions.

The progression we see is therefore from reactive → preventive → predictive → increasingly prescriptive maintenance.

The ultimate objective is not simply to predict that something might fail. It is to understand the condition, determine its likely impact and recommend or initiate the most appropriate action.

This can help manufacturers improve equipment availability, reduce unplanned downtime and make maintenance more efficient, while also reducing dependence on purely experience-based decision-making.

What technologies and architectures – such as edge computing, AI/ML, industrial IoT and real-time analytics – are needed to convert the enormous volume of sensor and vision data into actionable insights at the shop-floor level?

The most important principle is that technology should serve the manufacturing decisionnot the other way around.Manufacturing generates enormous amounts of data, but sending everything to the cloud and analysing it later is not always the right approach. Many shop-floor decisions need to happen within milliseconds or seconds, particularly when they involve machine control, quality intervention or equipment protection.

This is why edge computing is becoming increasingly important. Processing information close to the machine reduces latency and enables real-time responses. AI and machine learning can then be applied to identify patterns and anomalies, while IIoT technologies can connect equipment and make relevant information available across the manufacturing ecosystem.

OMRON has been developing this approach for many years, including AI capabilities integrated with machine controllers, real-time analytics and IIoT technologies. Our Sysmac platform, for example, enables manufacturing data to be collected from the shop floor and made available for IIoT applications.

At the same time, we need to recognise that IT and OT are increasingly converging. Manufacturing data should not remain isolated within individual machines or departments. It needs to flow securely and intelligently between the shop floor, engineering functions and higher-level business systems.

The architecture of the future factory will therefore be increasingly connected, distributed and intelligentwith some decisions being made at the edge and others supported by higher-level analytics.

The goal is simple: turn data into insight, and insight into timely action.

Beyond improving quality and reducing downtime, what measurable benefits can manufacturers expect from integrating sensing, machine vision and data fusion, and what are the key challenges they must overcome to achieve meaningful RoI?

The benefits can extend well beyond quality inspection and machine uptime. When sensing, vision and data are integrated effectively, manufacturers can improve overall equipment effectiveness, first-pass yield, throughput, changeover performance, traceability and resource utilisation. They can also reduce scrap, rework, unplanned downtime and the dependence on manual inspection.

But I would emphasise that the objective should not be to implement technology simply because it is available. The starting point should always be a clearly defined business problem.

Manufacturers need to ask: What are we trying to improve? Where is the current loss occurring? What data is required to understand the problem? And what action will the organisation take once that insight becomes available?

This is particularly important because one of the biggest challenges is not technology itself. It is integration. Many factories operate with equipment from different generations and multiple technology suppliers. Creating interoperability, ensuring data quality, maintaining cybersecurity and developing the necessary skills can be significant challenges.

Another important factor is scalability. A successful pilot should be capable of being replicated across machines, lines or plants without creating an increasingly complex technology environment.

OMRON's approach is based on integrating sensing, control and intelligence to address real manufacturing needs. More recently, our focus has also expanded towards data-centric factory automation and easier connectivity across existing equipment, recognising that manufacturers need practical pathways to digitalisation rather than always requiring large-scale transformation projects.

Ultimately, meaningful RoI comes when data leads to measurable action.

“The future factory will not be the one that collects the most data. It will be the one that uses data most effectively to improve decisions, empower people and continuously improve its operations.”

That is where we believe the combination of sensing, intelligence and automation can create sustainable value for manufacturers.

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