Updated: 2 days ago
Manufacturing organizations usually have a large number of production machines, the types of which depend on their activity. Variety is the rule: some are new, some are old, the brands often differ, etc. What unifies them is that they are all essential to the business and must operate 24/7. Every outage has a direct and visible impact on quality, customer satisfaction, revenue and ultimately profitability – therefore quality of maintenance is essential.
Monitoring and predictive maintenance reduce downtime and alleviate repair operations
To keep the machines in working order, several levels of activity could be undertaken, depending on the maturity of the organization:
Corrective maintenance: repairing what is broken when it is broken. To run it, you must have a qualified work force and a sufficient list of spare parts.
Preventive maintenance: carrying out maintenance operations on equipment in working order depending on either the frequency (change the oil every 3 months) or the time of use (change a transmission belt every 1000 hours) . The aim is to keep the machine in good health to limit the probability of failure; in addition, preventive maintenance can be performed outside of operating hours without affecting plant productivity.
Predictive maintenance: the idea is to detect signs of fatigue and deduce as precisely as possible when (and what) repair should take place, before the equipment breaks down and affects the production line.
A recent article by analyst firm McKinsey offers some estimates of the impact of well managed predictive maintenance, in the case of escalator maintenance. It shows an impressive reduction in the failure rate of 60% (repair is done just in time) and a significant improvement of 25% in the repair rate on the first visit.
Predictive maintenance requires collaboration
Even in this relatively simple and specialized example, we see interactions between:
Connected devices, usually sensors such as temperature, sound, vibration, torque, etc. that capture the health parameters of the equipment;
A collector (for example an IoT platform) which receives the data via different types of networks (LPWAN, Wifi, 3/4 / 5G…)
A specialized application, potentially based on machine learning to adapt to any real situation, able to transform all the raw data captured into actionable information such as the state of health of the machine, the type of potential fault and an estimate when a failure could occur. Depending on the solution, this application can work from cloud, edge, etc.
Last but not least, a maintenance and support team responsible for validate the diagnosis; decide which support operation should be undertaken, when and by whom; organize the repair according to the production schedule and inform the team in charge of it; and finally follow-up of the repair until the (prevented) incident is closed.
The efficiency of the predictive maintenance process depends directly on the quality of collaboration within the entire detection / decision / operation chain. Even in a close, siled deployment, it requires seamless interactions between connected devices, applications, and people from different teams.
However, in real life, isolated deployments are generally not applicable as systems, applications, and teams are multi-tasking and must share and reuse processes, tools and applications that are already available. Thus, systems like the predictive maintenance described above will need to interact with an even larger ecosystem, making the collaboration challenge even more critical.
Benefits of extended collaboration: the I.S. standpoint
Extended collaboration with the larger Information System could help make the maintenance solution more efficient and easier to operate by interacting with:
More sensors, connected on several platforms with different technologies, for better information collection;
Web services for better situational awareness, like weather forecast, pollution index, parcel delivery information;
Business applications such as inventory management and people calendars for better work organization.
Connecting maintenance tools to the rest of the information system, extended collaboration ensures better efficiency and simplified management at the plant level. In addition, it offers the flexibility to easily cope with the constantly changing processes and technical environment of today's industry.
Benefits of extended collaboration: the maintenance team standpoint
Digitization projects only work when people - employees, partners, citizens - embrace them because they are both simple and useful. Industrial maintenance involves:
The central maintenance department, in charge of operations - and ultimately of the positive return on investment of the entire process;
The people responsible for carrying out the repairs, whether they are employees, subcontractors or third party partners;
Local support teams, generally not specialized (in our example of an escalator, the technical support of the airport or the shopping center where it operates) but are essential to quickly apply preventive actions that protect the proper functioning of the devices.
Extended collaboration allows people involved in repair processes to communicate better, for example by using the company's collaborative tool (like MS Teams or Slack) for employees and more general social media (sms, Facebook Messenger, Whatsapp, Line) for external partners. Notifications and alarms are quickly received, work orders are better communicated. Capable of making requests and sending commands, repair teams are becoming more interactive. Finally, global traceability can be provided at the level of the overall process.
Extended collaboration connects individual digital projects to the enterprise organization as a whole
Just as aligning individuals does not make a team, a collection of individual IT projects does not build a flexible and manageable industrial information system. To ensure the best efficiency, employee buy-in and a strong return on investment, corrective, preventive and predictive maintenance projects will benefit from being connected to the larger organization they serve.
Would you like to go deeper into this point and discover how natural language can contribute to the success of your IoT projects, please find more information on our website, subscribe to our blog or simply contact us at firstname.lastname@example.org.