Investment in industrial IoT and Industry 4.0 comes in many forms. We are seeing impressive new greenfield projects driven from the C-suite where new factories are being built with native data access for IoT platforms. We are also seeing brownfield connectivity initiatives driven by CIOs, corporate IT or plant managers, where leadership makes a plant or organization-wide effort to connect existing assets (see this whitepaper for key considerations to keep in mind when connecting legacy assets). Whatever it may be, these initiatives lead with connectivity. When the organization implements a specific system, such as using machine learning techniques to tune manufacturing equipment settings, or to enable predictive maintenance on a high-value asset, the required connectivity is already there.
By submitting your personal information, you agree that TechTarget and its partners may contact you regarding relevant content, products and special offers.
Of course, there are factories that haven’t adopted these types of initiatives, but this doesn’t mean industrial IoT technologies aren’t being deployed in those factories. In addition to top-down investment, we’ve also noticed a new, bottoms-up trend. Many manufacturing/plant IT professionals and engineers are recognizing specific industrial IoT use cases and using low-cost software and hardware to tackle specific problems. In many cases they are using maker hardware, like Raspberry Pi or Arduino, and open source or low-cost software to connect to and visualize data from those assets.
Here are a few specific examples:
- An engineering group installed cameras throughout the factory, each connected to a Raspberry Pi. The Pi would store images of the product at different stages in production, which were then compared to a reference image of what the product should look like using image recognition software. Any differences were flagged for follow up by an operator.
- A plant IT professional installed Arduino devices on machines that needed monitoring. He converted the device data into MQTT, and used a simple web client on a low-cost monitor installed next to the machine for visualization. This served as a low-cost human-machine interface and enabled data access at the plant and enterprise level.
- A couple of engineers recreated a factory process in one of their basements to prove out that machine learning techniques could be used to tune extruder speed at a much higher resolution than had previously been possible. The technique was then deployed in the factory.
There are pros and cons to this type of bottoms-up or DIY industrial IoT approach. On the one hand, these are the type of individuals you want to have working in your organization. When they see an issue or an opportunity for improvement, they tackle it. They don’t wait to be told what to do, and they don’t let lack of budget or resource get in the way. At the risk of going overboard, they are the unsung heroes of industrial IoT — they have the knowledge and skills to both recognize opportunities and implement technologies, and the passion to see their projects through to completion, often in their spare time.
The drawback to the bottoms-up approach is that it has the potential to lack strategy and vision. It’s akin to building a house room by room but without a master design. Chances are it will be a hodgepodge of rooms rather than a cohesive, maintainable habitat. Creating an industrial IoT program in a piecemeal fashion can produce a collection of unique, independent technologies and not an integrated or coordinated program. The hidden cost here is maintenance, the hidden risk is security. A standardized design and approach will help reduce cost and mitigate the security risk, albeit at the risk of slowing down the discovery process.
Considering these factors, it’s far better to strike a happy medium and enable these individuals with enterprise tools that offer enough flexibility to cover many different use cases, but don’t rely on a disparate group for maintenance and support. This approach encourages these high-value individuals to continue to identify and solve problems, and ensures the systems they create will add value to the organization for years to come.
If you’ve used this bottoms-up or maker industrial IoT approach, what systems have you implemented and what tools do you rely on?
All IoT Agenda network contributors are responsible for the content and accuracy of their posts. Opinions are of the writers and do not necessarily convey the thoughts of IoT Agenda.