A Long-Term Approach to Industrial IoT is Crucial for Success
The concept of machines communicating with each other is not particularly new. As far back as the early 20th century, Machine-to-Machine (M2M) communication was present, as wired communication machines exchanged information through signaling. Computer networking brought in more sophisticated forms that found applications in SCADA, telemetry etc.
As data sensors, networks, storage, processing and integration become more advanced and affordable, IoT adoption in the value chain is set to skyrocket. Our new-world approach to the Industrial IoT is basically an attempt to give it a better structure, so that it is easier to use and more intuitive to understand.
The biggest benefit of Industrial IoT adoption is that it provides an end-to-end value chain view, rather than the traditional one in silos. This concept, called the Digital Thread, refers to a communication framework that can provide an integrated view of an asset throughout the manufacturing lifecycle, thereby enabling data-driven interventions and improvisations. Despite its amazing ability to drastically improve operations, the journey to Industrial IoT adoption needs to be well-planned and thought through.
Here are a few ways and some considerations to ensure that we get the most from our IoT investment.
1. Provide great experience with end-to-end data visibility
Shop floor managers today need to deal with multiple complexities that grow with every new product or innovation – from processing data from multiple sources to handling the demands of designers and engineers. They must screen information across various machine lines and other resources on the shop floor, and monitor the throughput from their shift. And they must do all this while ensuring minimal waste and downtime, low cost, and high performance on KPIs.
What they require is a system or a ‘system of systems’ that can give them end-to-end visibility into everything that’s happening in their department. This includes information on inflows into their work area, the way output is being applied to the next work area, dependency on customers and suppliers, and how their KPIs are getting impacted in real-time.
This empowers shop floor managers to drive efficiency and quality while reducing waste for the organization, to lower costs and shorten time-to-market.
2. Insights derived from data for action
For any IoT solution, the ability to acquire data is one of the core components. This data is then analyzed to help generate useful insights that translate into actions. IoT solutions can be designed to either recommend an action or initiate corrective action automatically. For example, data collected from chillers or heaters can be analyzed and designed to generate a service request automatically as part of preventive maintenance. Another example is a robotic arm that can not only spot a defective piece on the conveyer belt but also take remedial action by physically removing it.
Today’s IoT solutions can be designed so that the required action is triggered directly from a platform such as Infosys Nia or any other.
3. Innovation backed by data
Innovation is important because every industry constantly seeks to improve its top-line (through new products) and bottom-line (by cutting costs, improving throughput). An effective innovation process needs to be backed by data, rather than intuition.
Traditionally, the innovation process went through the following stages – market survey – design – engineer – manufacture. In this process, information on product performance from a design, engineering, quality and cost perspective was disjointed and tentative, and existed in silos, leading to higher rates of failure.
A combination of Product Lifecycle Management (PLM), IoT and Mechanics of Materials (MoM) can help improve product predictability and reduce the cost of innovation. We can predict product success and quality more accurately, thereby reducing recalls and quality issues in the market.
There are countless use cases for this across the manufacturing and services industries, limited only by imagination.
4. Accelerate processes
Whether traditional or digital, every organization wants to do things faster and reduce time to market for its products. The IoT process lends itself best to the Agile approach rather than the Waterfall model. The most effective IoT implementations are those that adopt approaches such as DevOps, Agile and Scrum. They also enable enterprises to work closely with customers and incorporate customer insights.
Also, it is important to ensure that the implementation team includes people who are well-acquainted with the industry and have a good contextual understanding of the problem to be solved. This is something that we follow at Infosys as well.
5. Ensure the right outcomes
In the past, industrial plant networks were very secure because they only consisted of proprietary protocols and communication technologies to connect various machines. Industrial automation control systems and protocols are now being transformed from proprietary to TCP/IP-based to enable Information Technology – Operations Technology (IT-OT) convergence to facilitate advanced analytics, remote operations, and other capabilities.
With the integration of IT-OT networks, plant networks are exposed to the same vulnerabilities and security challenges as the cyber world. Traditional cyber security frameworks cannot be used for the IIoT. The solution needs to include risk assessment of the overall system, down to the last machine/controller. At Infosys, we have built a framework for this called RAMS (Reliability, Availability, Maintainability and Safety). We need to implement trust networks/trust zones that are as safe as the old plant networks, using new risk assessment models. A good approach is to build in checks during the design and implementation phase to identify and flag risks and mitigate them either through technology or process interventions.
The Industrial IoT is set for massive adoption over the next few years. Yet there are some constraints that are holding organizations back. The first is the need for massive change management. The IoT can transform traditional business processes and roles of various people in the value chain. Organizations need to invest in training and show commitment right from the C-level to change mindsets and ensure comfort.
The other key concern is technology integration or interoperability, since data will come from different sources, following different standards. Ensuring that data is consumed effectively without getting hacked and without infringing patents is important.
Most organizations take a rather shortsighted approach to IoT adoption, looking for ROI quite early in the cycle. But true ROI can be realized only when the IoT is adopted from end-to-end in the supply chain to give visibility across the entire digital thread. Investing time and effort in PoCs and planning for effective change management can help organizations take the leap into the Industrial IoT. Patience and a long-term view can go a long way in improving their chances of success.