How to Capture Industrial IoT (IIoT) Value Through R&D Innovation
A guide for manufacturing enterprises looking to accelerate the pace of innovation
A guide for manufacturing enterprises looking to accelerate the pace of innovation
Use cases need to be digitally supported, but heterogeneous systems and application landscapes can become obstacles due to outdated software and/or decentralized solutions and technologies. These technical problems, combined with unclear business plan hurdles and lack of organizational capacity, end up driving many companies into a constant pilot purgatory.
In recent years, there have been significant advances in technology, especially in scalable connectivity and integration, which finally allowed enterprises to upgrade and extend their existing solutions rather than fully replace them. From 5G to AI to edge computing and beyond – the emerging technologies enable companies to implement and scale efficient use cases with minimal additional cost when used wisely.
Over the years of creating and delivering high-performance, customized IoT solutions, we at rinf.tech have noticed that manufacturing companies with a holistic approach are the most successful ones. In particular, they benefit from the digitization of industry along the entire value chain, providing a range of advantages to both their suppliers and customers.
This article focuses on the key technological changes enterprises need to make in order to support their innovation goals and how R&D can be an effective driver of IIoT success and sustainability.
Reaping the benefits of IIoT requires a holistic approach to drive the end-to-end transformation of an organization and technology.
To successfully develop IIoT projects, enterprises must focus on three core pillars:
There are several good reasons why companies should start or continue to use IIoT and advanced technologies:
To name just a few:
First, companies need to identify, prioritize, and test new use cases. They then need to create an achievable roadmap for deploying these use cases across IT, OT, and all enterprise/plant locations. After that, it would be great to capture value and measure opportunities with overarching impact and without distracting attention to local requirements for solutions. However, in terms of continually improving local initiatives, they need to be monitored and continued.
Organizations need to set clear targets for the entire digital transformation and create a team/unit responsible for monitoring progress and adjusting courses when needed. Then a new way of working should be introduced that fosters cross-functional communication and develops related skills and capabilities.
First, enterprises need to define the current situation and target architecture of their IIoT platform, focusing on collecting, connecting, ingesting, and integrating data in ways that enable innovative use cases, including platform cybersecurity management.
Second, you need to understand the impact of cloud computing and integrate it into the overall design of the platform.
Third, an ecosystem of suppliers and partners needs to be built to support implementation, taking into account the different complexity levels of individual enterprises. They include a type of production and products, enterprise size and IT-OT landscape, etc.
Research and development (R&D) is the part of a company that seeks knowledge to develop, design and improve its products, services, technologies or processes. Along with creating new products and adding features to old ones, R&D investments integrate different parts of a company’s strategy and business plans, such as marketing and cost reduction.
R&D consists of investigative actions that a person or business takes to obtain the desired discovery result. A completely new product, product category, or service will be created.
The development part refers to applying a new science or thinking so that a new and better product or service can begin to take shape.
Research and development is essentially the first step in developing a new product, but product development is not only research and development. As an offshoot of research and development, product development can refer to the entire product life cycle, from concept to sale, refurbishment, and retirement.
Deploying IIoT use cases is facilitated mainly by the following factors:
IoT and IIoT platforms become more user-friendly as the software supports user extension. In addition, as access to IIoT software development expands, more experimentation will take place to benefit the entire IIoT manufacturing ecosystem.
A rich set of ready-to-use APIs, code snippets and microservices, well-established communication standards, and available solutions with little or no coding ensure more cost-effective creation and implementation of use cases.
These low-code/no-code platforms offer small and midsize businesses access to IIoT software functionality without the need for talented programmers that early adopters should have had.
Routines for rapidly provisioning and updating applications in reusable data pipelines, containerization for hardware-independent software deployment across multiple peripherals, and sophisticated solutions to manage the entire software/hardware stack enable DevSecOps methodologies to be implemented and help scale platforms rapidly across the enterprise.
For enterprises migrating to Industry 4.0, 5G offers corresponding short-term IIoT opportunities. With these dramatic productivity gains, factories and plants can overcome the severe disruptions inherent in pre-Industry 4.0 production halls. As a result, manufacturing is expected to account for more than half of all 5G sales when 5G brings significant benefits.
Some of the most compelling examples of 5G in manufacturing are automated guided vehicles and real-time process control. Today, automated vehicles typically connect to a Wi-Fi network and rely on software loaded on the vehicle to manage routing and task execution. 5G will dramatically improve connectivity to the edge of the cloud for real-time sharing data between vehicles and coordinating fleets, thereby allowing them to free themselves from fixed paths in an ever-changing manufacturing environment.
As an added benefit, 5G also improves reliability by offering seamless connectivity as automated vehicles move through the factory and switch between access points or radios. When it comes to real-time process control, wired controllers and sensors can be used for years. However, 5G opens up a new space for solutions, helping to connect modules for relatively easy targeted analysis of sensors and actuators – especially important for older machines, whose control and power systems may be outdated and less adaptable to modern technology.
Artificial intelligence represents the third era of computing, usually defined as the ability of machines to perform cognitive functions as or better than humans. These functions include perception, learning, reasoning, problem solving, contextual understanding, inference and prediction, and creativity.
The convergence of groundbreaking research, business use cases, skyrocketing data growth, and improvements in computing power and storage are all encouraging advances in AI.
As there’s no single standard marking a clear distinction between weak AI and strong AI, both AI researchers and AI project managers find it difficult to make the right decision about architectures and technologies that should be used to deliver innovative AI-driven experiences. R&D provides a solid platform for experimentation, which results in better justification of corporate AI spending.
In particular, R&D allows for extensive experimentation and discovery around the following AI areas:
AI-powered decision making and processing happens closer to the data source generation, in contrast to the cloud, a technique known as “edge computation”. Internet of things and its billions devices combined with 5G network and increased processing power, has made large-scale AI possible at the edge of the possible.
Data processing directly on devices will be important in the future for healthcare, automotive and manufacturing applications as it is potentially faster and safer.
Microsoft designed HoloLens 2 head-up display (HUD) specifically for business solutions leveraging cloud and AI capabilities, interoperability with industry applications, and a suite of developer tools.
The device has already been adapted for the US Army and many of its advanced features include thermal imaging and night vision.
Smart glasses and headgear maker Nreal has announced an Android-compatible all-in-one headset for enterprises that looks more like a helmet than a pair of goggles, with built-in edge computing capabilities. These business-focused headsets are used in a wide variety of applications, from monitoring supply chains and complex equipment with digital twins, to conducting remote 3D meetings and providing augmented reality guides in on-the-job training.
These and other solutions have already motivated many R&D teams to experiment with AR, VR, Diminished Reality (DR) and integrate them with AI, machine learning, and Cloud.
Our rinf.tech R&D team is no exception. We’ve recently built a HoloLens-based digital twin of the robotic arm used for touchscreen testing.
Deploying early use cases is an investment, and therefore it is important to think about the deployment in a strategic manner. In most situations, it is recommended to deploy use cases based on value/feasibility.
First, select use cases that create immediate value and avoid use cases that require building databases from scratch or other complex prerequisites.
While the focus is usually on low-effort, high-value use cases, there is a strong case for using use cases with little or no ROI. For example, a use case may not result in immediate impact, but it may allow future use cases to be used.
This decision is pre-analyzed, and the rationale must be clearly stated so that all relevant stakeholders are aware of the investment and its value, even if that value is delayed.
Not all items need to be securely fastened before the deployment. In particular, even before the required data is connected, prioritization of use cases provides an overview of the required data sources and an understanding of what data is available and can be connected. The data collection architecture can be customized or defined for a specific OT equipment and data source based on this information.
However, in parallel, the development of the platform core should not be neglected and should start in sync with the development of the use cases. It is necessary to connect sources for automatic data collection as soon as possible. Depending on the implementation state of the target architecture, a use case can be deployed to a local server based on a specific architecture. This assumes the use of frameworks that will eventually be used on the back end.
But it won’t take long to implement the final architecture prioritizing the first use cases, and the migration of the first use cases will be possible in just a few weeks. Further, each use case is implemented in the target architecture. In preparation for this, consider and engage the next wave of use cases by connecting to the required data sources in the previous wave (s).
The point to keep in mind here is that many use cases take longer to prepare simply because some data sources do not yet exist. To compensate for this, it may be necessary to add sensors, change the PLC logic of the machine, update systems, and so on.
A standard process of connecting to new businesses should be established to ensure that the team is in place when deployments begin there. Implementation is also about changing how an organization works and thinks; thus, it is important to make informed change management the same priority as the deployment itself. The success of a use case can be undermined if the benefit has not been clearly identified and people are either not using a new application or system in their day-to-day work or are using it incorrectly.
A central IIoT deployment team takes care of the needs across the enterprise, making sure everything is in place prior to deployment. The list of needs includes the skills and the technical background required for a use case. The R&D team identifies the right people on-site to manage the deployment and resolves issues. The central team also monitors deployment status and emerging issues.
This provides a holistic view of benefits, implementation status and problem solutions to generate best practices for IIoT innovation.
A deeper look at use cases will almost certainly reveal a need for skills that may not currently be available in an organization or individual business. Hiring new talent or training existing employees should begin as early as possible so that skills are in place when they are needed.
Skill development and capacity building is an ongoing effort because deploying use cases never ends. Investing in these areas is a must for an organization that wants to succeed in its Industry 4.0 efforts.
It is also important to have a clear collaboration model that identifies one person for each workstream with clear accountability. Effective communication must be ensured within the organization, such as the exchange of information between enterprises and between the organization and its partners. Important archetypes of value chain partners are industrial equipment suppliers, technology providers and integrators, as well as specialists in functional areas.
Employee training should begin before deployment (or even concurrently) to ensure a shared vision and understanding of each use case capabilities and advantages. Training sessions are critical as time is devoted to answering employee questions and concerns. In addition, getting the people who will be working on a new application or tool into the development process and getting their feedback can create a tool that better suits their needs and provides broader acceptance and interest.
It will also be important to define the expectations and responsibilities of the use case owners as they will define the subsequent sprints within the use cases and report status updates and issues. Consistency of processes is very important, such as how communication occurs and measuring status and progress.
In addition to this, R&D teams can see future business opportunities. The end result of this should be research, prototyping, and clarification on the IoT product. The R&D team can study the idea before it becomes a product and support the transformation of the concept into a functional asset that solves the user’s problems.
IoT technology is evolving at a rapid pace. Why not save time and effort by testing ideas with a Proof of Concept (POC), Minimum Viable Product (MVP), or Prototype before launching a specific product?
POC is the practice of deep product interpretation. This happens before any planning for a product. POC allows you to learn a concept without spending too much time and money.
Prototype has the same basic goal of helping companies understand if a product is potentially good enough to get the job done.
The R&D team can test the product’s performance, design, and benefits by creating a prototype. This way, you can use the prototype with a limited group of people to test it out before going to market for multiple users.
MVP is a product with a sufficient number of features that can be implemented in the market. The goal of MVP is to provide feedback for the future development of IoT projects. Choosing between one or all of these three can be a little tricky, and doing one of them requires specialists and some IoT equipment.
Digital transformation and innovation projects often lead to overhead costs. This includes administrative costs in excess of the total cost of the equipment. Using R&D for IIoT projects will help reduce these costs. Many tech companies provide their R&D services to businesses looking for a variety of outsourcing options. Some of these companies even have entire departments that focus on experimenting with cutting edge technology. They investigate and scrutinize IoT solutions, deep learning, robotics, digital twin and other technologies to create solutions, POCs and MVPs for themselves and their customers.
The company must interpret risks and be able to assess or mitigate them immediately. Using R&D for IoT projects can help you mitigate these risks. R&D engineers need to understand well how new IoT devices or functions will work and whether they can meet your business goals and requirements.
Like any other technology, IoT solutions need deep research and development and are best done by professionals. R&D outsourcing allows you to optimize costs, get an idea to market faster and avoid life-saving risks.
Wrapping up, expanding market participation, cost management benefits, developing marketing capabilities and keeping up with trends are all reasons why companies invest in R&D. Research and development can help a company follow or stay ahead of market trends and keep the company up-to-date.
While R&D requires complex engineering-heavy resources, quality certifications, compliance with the industry standards and more, the innovations resulting from it may actually work to lower costs through more efficient manufacturing processes or more efficient and higher-quality products.
To accelerate enterprise innovation and create a unique and compelling IIoT value proposition, i’s highly recommended that manufacturing companies establish technology partnerships with companies that already have well-established R&D centers or innovation labs and that have accumulated best practices of leveraging the newest technologies and methodologies.
We'll share tips for choosing the right custom IoT software development partner in our next blog post. Meanwhile, check out what it takes to build a highly effective IoT development team in-house or offshore.