Assets in capital intensive industries are challenging to operate and even more so to maintain. The large number of parts to track, each requiring a similar effort to analyse (regardless of the value), means that the scope of any maintenance optimisation is narrowed. However, there are many more parts ranging in the $10ks than in the $100ks. That means the cost of not planning appropriately for these “less” expensive parts quickly adds up. Regardless of the individual part value, the key to unlocking savings is accurate analysis.
Maintenance transactional data records are typically entered by frontline workers, technicians and engineers, in remote and high-pressure environments. On a large mine site, 100s of notifications and work orders may be created daily. The accuracy and completeness of these records will vary significantly depending on the nature of the maintenance task and whether it was planned or unexpected. Reliability engineers typically need to validate and make corrections to the data before any analysis can be performed. Errors during this process are inevitable, because of the variable data quality and significant number of transactions to process. Next they need to identify a part’s behaviour in order to infer failure causes and optimise strategies.
Initiatives to improve data quality are often long, expensive (time and $) and unsustainable due to the number of human hours required to manually fetch, interpret and correct the data for every single part. These initiatives typically do not scale well, and due to the required data assumptions are inflexible to development. Due to these issues, few can afford to embark on such initiatives. Automation is a more appropriate solution that until now lacked the domain expertise to have a material impact.
IronMan® leverages intelligent algorithms and deep maintenance expertise from F1® and aerospace to ingest and validate transactional data in minutes so that reliability engineers can effortlessly scale their analysis to every single part. IronMan® is automated and currently processes over 10 million work orders around the world, every week.
Maintenance strategies can only be effective if the failure modes (premature, random, wear-out) have been correctly identified. Making that identification accurately is dependent on the interpretation of data with variable quality, and developing a strategy for the wrong failure mode can be costly.
IronMan® automates the ETL for all transactional data (including legacy ERP systems) so that reliability engineers are best equipped to accurately determine failure modes and optimise their maintenance strategies accordingly.
Maintenance plans schedule the replacement of parts after a certain age (calendar days, hours of use, fuel burnt, etc), and task lists detail the parts required to perform these maintenance tasks. Using standardised master data will reduce supply costs and maintenance planning errors. Replacing larger parts often involves replacing many smaller parts they are connected to, and sometimes the parts can be produced by different manufacturers. The end result is that if task lists are not optimised and maintained, it is quite common for parts that are essential to be missed, parts that are not required to be ordered anyway and discarded at site, and also for obsolete parts to be purchased instead of the latest and greatest.
Understanding the part ontology and identifying where waste is occurring in real maintenance activities, is an automated function of IronMan®. Every year, IronMan® prevents millions of dollars spent on unnecessary parts.
Unexpected failures lead to early replacements, sometimes done under direct purchase or with different part numbers, and hence the real lifetime of that part is reset. However, the ERP planned maintenance schedules are not aware of these replacements, which then leads to parts being replaced too early and wasted life. There are also clear benefits for mine production and safety if maintenance work is not done at all.
IronMan® automatically knows the history of every part, across all assets, at any time, and is therefore able to flag parts that do not need to be replaced in upcoming maintenance tasks. IronMan® has prevented the wastage of tens of millions of dollars in unnecessary preventative maintenance.
Warranties are normally only tracked in expensive parts, and there is much leakage of possible claims for all parts. This is simply because the time it takes for an engineer to verify if a part is eligible for a warranty claim is too great.
IronMan® knows the history of all parts, and is able to determine if its age makes it eligible for a warranty claim. To date, IronMan® has provided data to reclaim millions of dollars from OEMs.
Large companies typically try to operate the same OEM equipment and models across multiple sites around the world. Understanding how the maintenance work performed on that equipment varies, is a significant value driver and normally connected to standardised work method. As an example, a large train operator using the same type of locomotive across USA, would have many maintenance locations for performing work on those locomotives. For the same task (e.g. changing a bogie assembly), they are likely to be performed differently at each maintenance workshop.
IronMan® provides the visibility which is required to objectively monitor maintenance team performance across any number of sites, and automatically ranks them. IronMan® is typically able to show a 30% improvement in team productivity and on-time delivery across sites.
The overwhelming quantity of parts to maintain and variable quality data can explain many of the previously identified costs and inefficiencies. Most engineers know that their data is bad, and few are optimistic about extracting meaningful insights from it let alone with software.
IronMan® operates as software as a service (SaaS) and requires minimal training, because it was designed by engineers that are passionate about improving maintenance. If you would like to extract actionable insights and savings from your transactional data, now.
© Ox Mountain Limited 2020