2015 in review: the top 10 lessons from Industry 4.0
In 2015, more blog posts were devoted to manufacturing, and Industry 4.0 specifically, than any other subject. On average, these posts also met with the greatest interest among our readers. Industry 4.0 was a motif running through all topics the whole year long. I’d like to look back at what happened in 2015 from an Industry 4.0 standpoint and summarize where we are now. And now, my personal Top 10 list of Industry 4.0 lessons in 2015:
Lesson #1: Industry 4.0 needs value stream organizations: how far along are companies in 2015?
When manufacturing companies in Germany undertake continuous improvement projects, they focus heavily on their value streams. This is in essence a perfect fit for Industry 4.0, as it requires connectivity across boundaries: a value stream by definition represents the flow of parts from supplier to production to the customer. However, the primary focus of many projects in 2015 – whether for lean production in general or for Industry 4.0 specifically – was on the benefits produced by optimizing certain individual processes. If you’re looking for total, all-encompassing connectivity – covering all links in the supply chain and reaching out across multiple plants – you’ll still have to wait a few years. Over the next couple of years, organizations will increasingly (have to) ensure that their structure reflects their value streams ; this means moving towards value stream organization. This does notmean, however, that Industry 4.0 offered no optimization potential in 2015. Quite the opposite: for companies that reviewed their value streams specifically to pinpoint where problems occur and how these can be improved or resolved with software, there were immediate benefits to be had.
Lesson #2: Condition monitoring and predictive maintenance: where do Germany’s manufacturers stand in 2015?
In 2015, condition monitoring was named most frequently as the application that production experts truly need. Predictive maintenance is not far behind – at least, that’s the plan of German production experts. I’m curious to see if we’ll be talking about this in 2016 as one of the blog’s topics of the year.
Condition monitoring generally refers to solutions that collect, analyze, and monitor machine, process, or quality data. Cross-machine connectivity of data is the crucial prerequisite for making machine status transparent and in turn for taking action (e.g. having service technicians recalibrate the parameters) if values fall outside an acceptable range. See our cycle time monitoring example to learn how it pays off.
The analysis incorporates data from MES, data sent directly from the machines, data from sensors, and – depending on how automated the derived actions are – it might also include data from ERP systems, for example.
People often ask us: so what’s new about any of that? Here’s the answer in full detail:
- The fact that connected systems are the data sources for analysis.
- The way results are presented visually and in a central monitoring view. This is often the first time that experts have had convenient access to – and an overview and direct control of – more than just individual machines (e.g. an entire production line).
- Only the right information is passed on to the right people at the right time via the right medium.
- Experts can now take their knowledge of production, maintenance, repairs, etc. and feed it directly into the software solutions. They can use existing data and define rules on when to react to certain conditions – and don’t need to wait for IT experts to be available for implementation.
- These systems have to be scalable (start small, think big) and interoperable (open interfaces, compliant with existing Industry 4.0 standards, integrated into the existing IT landscape of the enterprise).
Lesson #3: The biggest Industry 4.0 challenge in 2015: how to get started
For many companies, the biggest challenge of 2015 was to identify Industry 4.0 projects that can start delivering benefits and providing a competitive edge today – which meant fitting themselves into an Industry 4.0 “big picture” that is only just starting to take shape at each company. What’s needed here? Three things are essential (see also the Industry 4.0 Innovation Cycle) – but none of it is rocket science.
- To get started, companies need to equip their production experts with the right tools. These include rules technology for their business experts and big data processing, e.g. using data analytics tools, so they are in a position to exploit the huge quantities of existing production data.
- The infrastructure setup has to be “Industry 4.0-ready.” This includes connected machines, a data collection infrastructure, suitable security mechanisms, etc.
- Last but not least, your organization should already have a rough sketch or initial concept of what Industry 4.0 means for your business.
The best way for you to learn “the rest” is in an actual project! Our experience from over 100 Industry 4.0 projects at Bosch shows that companies achieve the greatest success when they view Industry 4.0 as an evolutionary development ; in other words, when it unfolds step by step. A retail and logistics company might for example start with a PoC in which a small number of connected sensors (temperature) are deployed to monitor the cold chain. The goal of this pilot phase is to generate measurable quick wins. Once this phase has been completed successfully, the next step is scaling. In this phase, the number of sensors is massively increased; now thousands of them monitor the cold chain. The company can integrate more sensor data (vibrations, geolocation) and develop completely new applications. In further phases, this sensor data can be put to other uses, for instance to optimize the ordering processes. This step-by-step approach is well-suited to small and medium enterprises (SME), too.
Since we usually talk about new Industry 4.0 applications for which there are currently no off-the-shelf solutions, it is essential to choose the right partners: ones that offer pooled expertise with which to develop these sorts of new solutions. This expertise primarily includes know-how in plant engineering, product development, production processes, sensor technology, data analytics, software solutions and integration, UX, legal issues concerning the IoT, and sometimes even more.
Lesson #4: What was the biggest obstacle in 2015 that kept companies from moving toward Industry 4.0?
This was a question posed by Bosch Software Innovations when it conducted a survey in summer 2015 among 180 production experts in Germany, Austria, and Switzerland. According to the survey results, the biggest challenges concern
- Experts: Oftentimes there are simply not enough experts with the necessary combination of skills. This is particularly true for the combination of manufacturing and IT expertise.
- Standards: Many companies are waiting for standards to be instituted. The fear is that they will make the wrong investment decisions today and then have to start from scratch again tomorrow.
- Security: “Production data in the cloud” – this idea sets off alarm bells for the majority of production experts. Public cloud, private cloud or no cloud at all? How should I be conducting risk and threat analyses? What security mechanisms do I need to use? And what else should I do to secure systems and machines, and protect our intellectual property – not just to preserve know-how that sets us apart from competitors, but also to keep it safe from manipulation?
- Benchmarks: There is a need to establish benchmarks of finalized projects. The insights (e.g. calculations of RoI) offered by such benchmarks can be helpful when launching an in-house Industry 4.0 project that is intended to yield the greatest possible benefit in the core business. This can often be something simple, like a practical device that provides production experts with information relevant to their work.
Lesson #5: Security in Industry 4.0: what has to be done?
There’s a good chance that “production data security” will become the topic of the year in 2016. Without a doubt, security is crucial to the success of Industry 4.0.
Security by design has long been an established principle in software engineering. Secure engineering processes have always included risk and threat analyses as an essential part of software development.
What’s new here is the IoT context. An Industry 4.0 IT setup involves any number of stakeholders, systems, and other factors. Moreover, because these systems evolve, you need to be prepared for scenarios you can’t even envision yet. The best way to cope with this is to find partners who specialize in the IoT: partners that are involved in the development of appropriate IT security concepts, architectures, and standards and can help you establish these in your organization.
Suitable partners will also be familiar with software architectures in manufacturing from years of practical experience. For example, if you’re looking to implement remote services, the necessary infrastructures are already available for you to build on. What’s more, many initiatives that are now just getting off the ground are tackling this topic in full. Over the next few years, they will create dedicated blueprints, reference architectures, best practices, and more for security in Industry 4.0.
Bottom line: ensuring the integrity and security of your data should be your top priority.
In my first blog about the 10 lessons that I take away for Industry 4.0 in 2015, I focused on the following five topics: Value stream organizations, condition monitoring, how companies can best get started with Industry 4.0, which roadblocks they see and the role that security concerns play in Industry 4.0. Now there come the next five of my top 10 lessons.
Lesson #6: Generating knowledge and added value from data is the essence of Industry 4.0
Since at least 2015, if not earlier, it’s been clear how important this topic is for production: data is the oil of the connected factory. One of the core competencies in Industry 4.0 is the ability to transform collected data into new information, consolidate this into knowledge, and then incorporate it into production management; in other words, reap new benefits. Companies in Germany are moving towards analyzing data at several points along the production process and gathering this new information. Doing so allows them to, for example, improve production cycle times and detect deviations early on to avoid unplanned machine downtimes. The quantities of data associated with these activities is certain to continue growing.
To implement this approach, we need experts who are equally at home with data analysis, manufacturing and software engineering. It’s the combination of skills that’s essential. What do we mean by that? Allow me to explain using an example: as a rule, data scientists look at a problem from a data-centric perspective. Their approach is neutral and flows from the data. Production experts, on the other hand, focus on the technical side of the problem. Merging these two methods is the key to the art of efficiently resolving the problem using data analytics. Particularly in the business understanding phase, experts need the right kind of production know-how, yet shouldn’t lose their neutral perspective on the data to be analyzed.
Several blog posts have provided examples of the benefits resulting from such analytics projects in manufacturing. Based on the analysis of data from 30,000 hydraulic valves, Bosch was able to cut the time spent on end-of-line testing of these valves by more than 17%. Data analytics showed that, provided the results of several earlier steps were positive, certain subsequent testing steps in the inspection process could be eliminated. The outcome of those subsequent steps could be reliably predicted by analyzing the earlier steps.
Lesson #7: Gearing up to provide new services: the foundations have been laid
Discussions about Industry 4.0 frequently mention new services and business models. However, as of 2015, few of these had made it into actual practice.
One example of an Industry 4.0 service in production is Bosch Rexroth tightening systems, which come with a software-based cockpit for the intelligent analysis of safety-relevant tightening data. This extra software enables the supplier to sell an innovative additional service that sets it apart from the competition. Users benefit from the solution’s ability to detect faults in production early on so they can be avoided. As important as the software-based service itself is the fact that it comes with an open interface. The implication: data from different tightening systems can be integrated and visualized within the cockpit.
New machine maintenance services are made possible through a combination of configurable production rules and open PLC interfaces for direct access to the machines’ control units. This means service technicians’ know-how can be integrated into predictive maintenance services. For example, service technicians know that changes in engine torque indicate mechanical wear and tear. This knowledge can be translated into a rule that the software then executes: if an engine’s torque changes by a given amount, the system notifies the maintenance group accordingly. The constant rule-based monitoring of the operating data means that service intervals can be adjusted dynamically to accommodate the actual wear and tear. As a result, suppliers can organize their maintenance services more efficiently, and users benefit from increased machine availability and with it, better plant productivity.
Lesson #8: “The earlier, the better”: leveraging the rule of ten in production for Industry 4.0
Liability claims were certainly a cost-intensive issue for manufacturing companies in 2015, but not the only one. Quality problems can surface before the product has even left the factory, and the further along in the value chain a quality problem is detected, the higher the company’s costs. The relationship between the two is called the “rule of 10”. At each step of the value chain, fixing a problem is approximately 10 times faster and cheaper than doing so in the next phase.
Industry 4.0 software can have a huge effect on reaction times when it comes to process quality management. Instead of waiting for a quality report from somewhere along the value chain before identifying and solving problems in production, process quality is checked every time the process is run. If a quality problem is detected, troubleshooting can be started instantly.
You can use this principle to meet your zero defects quality targets. Now, with the increasing virtualization of production landscapes, you can collect data at any point in time from any location and analyze it accordingly.
Lesson #9: Nothing in Industry 4.0 is possible without collaboration
As Industry 4.0 is developed further, much will depend on the collaboration among companies, universities, associations, consortia, and standardization organizations. Many of the initiatives in 2015 were geared towards developing standards for Industry 4.0. I’ve provided some representative examples below.
Partners are collaborating on new solutions for various Industry 4.0 issues and making use of IIC (Industrial Internet Consortium) testbeds. One of these solutions is for an aircraft landing gear predictive maintenance use case.
A range of Eclipse communities are working collectively in open source projects on services relevant for the IoT and Industry 4.0, e.g. an IoT software update service and an IoT information model repository.
Other major cross-sector collaborations (particularly in Germany) include IUNO, which develops blueprints for secure Industry 4.0 projects, and Plattform Industrie 4.0. One of the latter’s goals is to draw up recommendations and initiate suitable standards.
There are also plenty of bilateral partnerships that focus on IoT and Industry 4.0 topics. For example, PTC and Bosch Software Innovations have teamed up to efficiently develop IoT and Industry 4.0 applications and make the management of machine connectivity secure, flexible, and transparent.
Lesson #10: Last but not least: what Industry 4.0 is not
Looking back at 2015, you could get the impression that all manufacturing activities carry the Industry 4.0 label. But now that I’ve summarized the top Industry 4.0 topics, allow me to list a few facts that make it clear what Industry 4.0 is not:
0.00 avg. rating (0% score) - 0 votes
- Despite the concerns of the public or labor representatives, analyzing personal employee data is not at all a target of Industry 4.0 projects, according to manufacturing experts in Germany.
- Also, Industry 4.0 is not about reducing labor costs; instead, it’s about creating new working conditions, new career profiles, and training opportunities. Ultimately, all Industry 4.0 projects aim to improve competitiveness and in turn secure jobs (especially in countries with high labour costs).
- It’s about more than just lean production activities that have been in place for many, many years in our highly optimized manufacturing plants.
- It doesn’t necessarily involve “the cloud.” Condition monitoring can rely on data from connected machines without it having to be managed in the cloud.
- It’s not a technical topic. It’s a business topic, just as other IT projects are – with software, hardware, sensors, and more technology being the enablers.
- It’s not just about digitalizing production. Industry 3.0 was the information technology phase, and Industry 4.0 is about leveraging connectivity – and the benefits that result from analyzing connected data.
15/01/2016 / arnaudportet / 0