What if you were able to predict the future? What if you were able to detect possible system defects and fix them before they resulted in failure? What if you were able to anticipate possible accidents in order to prevent them? What if you were able to predict the lifetime of your hardware and thus be able to replace it seamlessly?
Well, you can! With the development of IoT, cloud and machine learning, predictive maintenance systems are on the rise, making the collection, storage and processing of industrial hardware data increasingly simple.
Predictive maintenance uses IoT condition-monitoring tools and techniques to track the performance of equipment during idle, normal, and peak operation to detect and prevent issues. Data is collected using sensors and centralized in the cloud for analysis; predictive models are being designed to recognize warning signs or anomalies based on a history of failures; and a machine learning system is set-up to "learn" to recognize and give warning of new events and potential failures before they occur. These 3 steps allow maintenance supervisors to ensure the unfailing performance of the equipment whilst optimizing maintenance frequency to avoid unnecessary costs.
A recent study by Deloitte states that on average, the implementation of predictive maintenance strategies has allowed industrial enterprises to increase their productivity by 25%. How?
First of all, unplanned maintenance can lead to substantial downtimes and jeopardize productivity: teams that are interdependent can’t function, production is delayed, services are interrupted… By monitoring and analyzing the health of the assets on a regular basis, predictive maintenance can help to uncover issues and repair or replace them ahead of failure thus minimizing the production hours lost to maintenance.
Maintaining machines overnight on site is also far easier than having to repair a machine during business hours, especially if the machine in question is moving or transporting users. Alstom Transport —one of the leading integrated transport systems’ companies— in their quest to optimize the trade-off between system availability and maintenance productivity, have developed predictive maintenance solutions to tracks their trains. Their R&D department have designed IoT solutions that offer real-time monitoring of crucial elements in trains (rolling stock rotating elements). Information is fed to a hub then used by maintenance teams to better plan their interventions, at times that are convenient to all, which minimizes the time the equipment is offline. Their innovative and patented HealthHub™, offers a new approach to asset management through condition-based maintenance, providing greater efficiency in the overall maintenance process. Not only does the system allow them to save up to 20% in preventive maintenance labor, but they also save 15% in materials consumption by extending the remaining useful life of all types of rail equipment.
A recent study by Aberdeen Group found that the average cost of unplanned machine downtime is $260,000 per hour and the price has skyrocketed over the last few years. And that’s not just down to lost productivity! The failure of a single component can damage the whole system, impacting the quality of the production, creating accidents and endangering staff or customers. Ultimately, it could result in the loss of unsatisfied clients, or in a worse-case scenario, even threaten the very existence of the company.
Predictive Maintenance, with the help of monitoring devices, gathers a wealth of process data and advanced analytical methods to predict failures before action has to be taken. This allows for a better estimation of the remaining lifetime of assets.
In industrial plants, especially in the chemical and oil and gas industry, rigor is key. Well aware of the stakes, Total launched a project in 2013 that aims to provide remote monitoring of its entire fleet of rotating machines distributed at its production sites around the world. Several solutions capable of predicting anomalies in equipment have been implemented to remotely monitor machines: various sensors fitted to the machines can measure data such as oil pressure, oil levels, bearing, gas or water temperatures, or vibrations on the shaft lines. Nowadays, new types of retrofitted sensors and autonomous power solutions enable predictive maintenance on almost any asset or machine by connecting measuring instruments. One such Vibration sensor, the Sushi Sensor created by Yokogawa Electric, is being equipped with Saft’s LS batteries. Explosion proof, the device measures the vibration and surface temperature of equipment, a good parameter to identify the deterioration of motors, pumps, and other equipment as the vibration increases when an instrument deteriorates. Data is being fed in real time to a platform that provides advanced analytics thanks to AI and machine learning: the status of the entire equipment, its soundness, the trends, etc. It can precisely identify signs of abnormality and when an unusual behavior is detected by simple diagnosis, a detailed examination is performed to identify the cause and obtain sufficient evidence to judge whether maintenance is required and when it should be carried out.
Another example of IoT adoption in Oil & Gas facilities is the connected pressure gauge manufactured by Baumer, which measures the pressure in the double bottom of oil tanks (in which the vacuum is maintained). When the pressure increases an alert of possible risk of leakage is given and inform the maintenance crew that an intervention could be required. The use of predictive analysis gives Total’s plants’ both the reliability they need to ensure complete safety, minimal environmental impact and enhanced cost control.
These types of solutions have allowed some companies to reduce their breakdowns by 70% and increase their equipment uptime and availability by 10-20%.
From product supplier to service supplier, predictive maintenance can also benefit and bring value to your end customers. By implementing a predictive maintenance strategy, you differentiate your products from the competition which can become a real competitive advantage when it comes to signing contracts. Predictive maintenance can create and capture a new added value: the data collected by sensors placed on machines makes it possible to better manage the lifecycle of spare parts and to replace them before they break down, thus saving the customer a waiting time and boosting their satisfaction. The company then benefits from greater responsiveness, an improved offer that truly meets the needs of its clients, while building a trusted brand image. In short, a competitive advantage!
Bosch has created predictive maintenance solutions for the car industry. Thanks to their 360° predictive maintenance system, fleet operators can help prevent unexpected breakdowns on their fleet by predicting system and component failures. Sensors installed on all the components of the car (transmission, brakes, generator, exhaust pipe, oil pump, engine, belt and chains, etc.) continuously monitor the condition of the car and transmit the information to the control unit. An analysis software generates a precise status evaluation on a central server. The software calculates the probable remaining lifetime and forwards this information to the manufacturer who can then schedule maintenance and even send messages to the driver.
For the fleet manager, the prevention of breakdowns means avoiding downtime and unecessary costs, saving up to several hundred euros per vehicle and year in terms of material and personnel costs.
For the fleet management company, these two arguments are competitive advantages that can make or break a sale.
If you’ve read the article this far, it’s probably no longer necessary to explain that the ultimate goal of predictive maintenance is to save money! Better productivity, less breakdowns, an increase in equipment uptime added to costs savings in material spend and a better use of maintenance labor has resulted in—for the industries that embraced predictive maintenance—an average reduction of global maintenance costs by 25%
The Marine Division of Caterpillar operates tug boats and shipping vessels. A few years back, they installed shipboard sensors on their vessels to monitor generators, engines, air conditioning systems, GPS, fuel meters and more. They realized—thanks to the data collected from the ship-board sensors— that there was a correlation between the hulls’ cleaning frequency and the performance of the boats fuel consumption. Inefficiencies due to dirty hulls was costing them up to $5 million every year. More frequent cleaning - twice a year rather than every two years - would certainly increase their hull-cleaning budget, but the savings made on fuel consumption would surpass the expenses. Indeed, nearly $400,000 per ship could be saved by switching to an optimized cleaning schedule. That’s $20,000,000 per year for a fleet of 50 ships...
Manufacturers can even go beyond that by providing predictive maintenance services to their customers, an opportunity for new, recurring revenue streams. Depending on the type of activity and maintenance required, they can capitalize on the devices developed to offer their services to clients: analytics-driven services, dashboards, optimized maintenance schedules, or a technician dispatch service before parts need replacement.
Alstom Transport has thereby become the world’s no. 3 maintenance provider and offers their services on both Alstom and non-Alstom equipment. They provide three adapted service levels: technical support and spare supply agreement, a core maintenance contract with fully outsourced maintenance operation, and a full maintenance contract where Alstom become the sole fleet maintainer. As an example, in 2018 Alstom signed an eight-year full maintenance contract with Arriva for the maintenance of their Coradia Nordic regional trains. The contract was worth approximately 135 million.
Predictive maintenance would be practically impossible without IoT solutions capable of gathering and rapidly analyzing large amounts of data. And the advent of new technologies such as retrofitted wireless sensors, coupled to Saft’s long service life, reliable batteries, and low power networks, are making it increasingly easy to deploy surveillance solutions. But for these to be efficient, manufacturers need to have high reliability equipment, which is not necessarily the case for all IoT applications where sometimes the determining criterion is cost.
Reliability, Responsiveness, Revenues, 3 R’s that can make the success of a company and definitely add value to any business. The question now is to carefully choose which problematic needs solving with these new tools and to prioritize them….
Act today to prepare tomorrow. Get a full picture of your entire demand and supply process, Identify risks & bottlenecks and better manage your inventory
A quick recap of the webinar recently hosted by EMnify where Saft helped shed some light on how to choose the right battery for your Internet of Things connected objects.
Car parks are at the heart of the challenges of new mobility and Smart cities. So what does the future hold for this space in the city of tomorrow?
What is the impact of the various communication protocols on your application’s power consumption?
IoT has been instrumental in reducing cities' ecological footprint. Discover how innovative approaches to sustainable living have been developed with the help of new technologies.
Here are a few facts, figures and trends, gathered across the web, that can help us understand how the IoT world is shaping up and what to expect over the next few years.
Saft's cylindrical primary lithium cells are helping Techem's #IoT #SmartMeters provide value to residents by monitoring gas and water consumption. Here's how...
If IoT technology can’t eradicate viruses, it can at least help containing it and make healthcare more accessible.
Advances in IoT technology are helping experts to track and understand the entire journey of a species in its natural environment.
How do you track a large fleet of tank containers in a variable and sometimes potentially explosive atmosphere, in order to reduce operational costs with real-time data?