Optimisation of machine alarms: here's how you can quickly and easily make a start in connected production!

Are you interested in increasing the availability of your production machines and at the same time creating a solid basis for digitalising and intelligently networking your production?

Optimise your machine alarm management and at the same time open the way to 'data-driven manufacturing optimisation'. Our solution automates alarms, provides them with an intelligent set of rules, and in this way it also unburdens your service personnel. The architecture enables you to continuously expand the way you evaluate your production data.

Real challenges in production & manufacturing

Rule of thumb: a 'good' overall equipment effectiveness (OEE)  for a (German) manufacturing company is about 60% on average. Putting it simply, out of 100 products that could be produced under theoretically optimal conditions, in actual real-world operation just 60 are produced. Around 85% is regarded as an excellent OEE (average value from a study ). With our solution, we start off in 'light-weight' mode i.e. without major entry hurdles, because just optimising the flow of machine alarms can quickly represent a high economic value for companies without requiring large investments. Please ask us about reference examples.

To increase machine availability, we must first look at the causes of low equipment effectiveness.

They are often due to machine-related failures and faulty processes.

Few companies have effective information and warning mechanisms that signal plant malfunctions or necessary maintenance work in real time. Differing country standards and growing requirements in the area of obligatory documentation lead to increased efforts in quality assurance and thus to slower reaction times.

On the other hand, smooth production processes are often based on the know-how and many years of experience of a few experts in the production environment. In the whole global arena, however, there is often a shortage of qualified service personnel in the areas of machine attendance and operation. A pool of widely differing machines and increasingly stringent requirements in these areas, together with the lack of a holistic view in the event of simultaneously occurring process anomalies, make the situation even more difficult. Documented rules and procedures with corresponding prioritisations are often not immediately available when a machine failure occurs.

This results in high skills requirements with respect to fault prioritisation and correction, to maintenance and to operation. Incorrect decisions in prioritising the troubleshooting lead, in turn, to lost efficiency, extended downtime, defective products, unfulfilled quality expectations, and rejects.

Compliance with production standards is often made more difficult if these are not digitalised into the normal operation, but are only available as printed checklists, for example.

How can these existing potentials be leveraged by a networked and data-driven production system?

In launching suitable projects to eliminate the above-mentioned challenges, there is often a certain inhibition or reluctance on the part of the company. This is due to the anticipated complexity and the need to integrate numerous departments ranging from supply chain to production, quality assurance, maintenance and logistics.

With our approach, we systematically lower this inhibition threshold: our Proof of Concept serves as a lighthouse, which is initially implemented in a delimited area  and focused on the Use Case 'machine alarms' (our proposal is a link-up of 5 lines in a plant, each line having 2 machines). This approach can be extended as required in a step-by-step, controlled procedure. Depending on the company's goals and the project scope, the scaling can either take place across several lines and/or plants or include additional Use Cases as required.
 

With this solution:

• you do not lose the overview – not even in critical situations – thanks to prioritisation of alarms and filtering
• workers and operators are directed systematically to the causes of errors
• the availability of your machines is improved
• the OEE (overall equipment effectiveness) rises, thanks to reduced response times and a structured, predefined approach
• the onboarding of new employees is faster
• process reliability is improved thanks to compliance with, and digital documentation of, standard processes

Who would benefit from optimising machine alarms as part of a connected production system?

The solution+project presented here will be worthwhile for companies in the discrete manufacturing sector that meet at least two of the following criteria:

• So far, they have only a low level of networking
• They have differing types of machines, with very different operator interfaces
• They have a (globally) distributed pool of machines
• They suffer from the shortage of skilled workers in production and maintenance

Our technical solution and its advantages

Together with our partner, the database specialist Crate.io, we offer you a solution consisting of four modules:

1. Collection and normalisation of heterogeneous data. As required, additional enrichment of IoT sensor data with ERP data for simple Use Case processing
2. Digitalisation of standard production processes (Standard Operation Procedures – SOPs) from checklists via data entry and information output to 'Gemba Walks' (also 'go to Gemba', Lean Production approach)
3. Implementation of a fully automated notification system, with information relevant to rapid troubleshooting and compliance with SOPs based on a rules engine, including a mobile application and chat bot with voice output for detailing the source of the problem and the instructions for solving it.
4. Development of a comprehensive, intelligent logic for real-time anomaly detection covering all processes and machines that are included

Start now, scale later: our project approach for the start of connected production

On the matter of machine alarms, we recommend a lean, multi-stage procedure in a lighthouse-project approach with clear and measurable results. The delimited and multi-stage procedure not only reduces complexity but, thanks to the iterative approach, also helps companies to get a quicker and more flexible picture of the success of the networked-production project. If necessary, they can 'readjust' at an early stage. A further advantage: a comparable lighthouse project can often be implemented in three months without having to dispense with global production standards in the preliminary stages.

From our experience  we suggest the following phases.

The scope of the project is planned in Phase 1, the 'Visioning and Scoping' phase, which brings all stakeholders to the table. Suitable starting points are usually individual production lines or specific machine groups.

In phase 2, 'Connectivity and Data Platform', the architectural and technological foundations – for the topic of networking, for example – are created, and the system boundaries are jointly defined in expert discussions and workshops. The system architecture and IoT system design is defined at the machine or production level.

The last phase, 'Process Modelling and Go-Live', is ultimately about the analysis of the alarm management and its implementation and configuration in the defined pilot environment for the lighthouse project.
After the operator team has been briefed, the trial operation can be started. This will provide important insights for a possible Scale Out into the customer's organisation.

To optimise the management of machine alarms, we have identified the following potential benefits based on reference values:
 

In the 'Operation' area, the overall plant efficiency is increased thanks to faster reaction to machine downtimes resulting from the automatic and prioritised forwarding of alarms in real time – directly to the (mobile) terminal device. Abnormalities in data are detected at an early stage, which also leads to a faster reaction to quality problems, or even to their complete prevention. Reliability is improved thanks to compliance with, and digital documentation of, standard processes. The system for notifying the responsible persons and providing the information relevant to rapid troubleshooting and compliance with standard operation procedures is fully automated (a synthetic voice details the source of the problem and the instructions for solving it). This results in potential savings of up to 20% in personnel costs, because onboarding and training are simplified. Further potential can be created on a company-specific basis by specifying standard processes.

Customised consulting and implementation services from Zühlke, in conjunction with highly scalable databases such as Crate.io, offer the necessary flexibility to gradually, bit by bit, build up connected production and manufacturing. In this way, the knowledge gained can be rolled out to other systems to facilitate step-by-step connecting of your entire production.