News and Announcement

IBM Maximo Application Suite: Accelerated Asset Performance Management

From Model T to the IBM MAS-ters

Artificial intelligence can contribute significantly to prolonging the productive lifespan of assets. The key to achieving this optimisation for a firm requires three critical attributes: 

  • Asset and maintenance management aligned to the organisation’s corporate strategy and objectives 
  • The right analytical engineering skills to reap the benefits of a data-informed predictive asset management and investment planning system; and 
  • Comprehensive multi-layer Asset Performance Management (APM) and AIP (Asset Investment Planning) applications to simplify this process and augmenting your current Enterprise Asset Management systems 

Tony Turner, Projects and Innovations Director at Vetasi, reckons the new IBM MAS (Maximo Application Suite) is one of the most comprehensive APM applications available.

“The original Maximo transformed from a Computerised Maintenance Management System (CMMS) in the 1990’s to an industry aligned Enterprise Asset Management (EAM) system post 2000. MAS now extends on the EAM paradigm and enables the move towards APM and asset life cycle and investment planning.”

At Vetasi we call it “EAM Extended”. 

Download the IBM Maximo Application Suite Overview document to get more information 

Vetasi, a global asset management consultancy and part of Cohesive, is the largest IBM Maximo partner in Europe, Africa and the ASEAN bloc. 

Advantages of predictive maintenance 

Damien de Gouveia, a lead consultant at Vetasi, says the versatility of MAS is such that it can assist a firm in making the transition from prescriptive and condition-based maintenance modes to predictive maintenance whilst unlocking value at several levels in the process. 

Figure 1 An indication of both the maintenance interventions as well as maintenance expenditures of the various modes of asset maintenance.

De Gouveia reckons the advantages of predictive maintenance that is well enabled by MAS, are manifold:   

  • It allows predictions on the remaining useful life (RUL) of assets, which is a distinct advantage over planning using the assumed life cycle by the original equipment manufacturer (OEM). For example, if OEM indicates that a pump should be replaced after 10 years and the digital Big Date data flows of readings of temperature, volume flow, vibration and metal fatigue indicate that it is “good for a further 8 years”, replacement is unnecessary. The return on the capital investment in that pump increases tremendously. The benefit: optimal financial and productive returns on assets. 
  • It reduces the disruption of the production process: in the reactive mode where maintenance and repairs were done when the asset failed, costly production disruptions and unplanned maintenance occurred. In the planned (scheduled) mode of preventative maintenance (for example servicing a car every year or 15 000 km, whichever comes first) production is interrupted in a planned way, but the cost of maintenance is high since parts are replaced in the majority of cases whilst there is still a substantial RUL of assets. The benefits: less disruptions of the production process and lower maintenance cost. 
  • Condition-based maintenance that introduced predictive maintenance is coupled in MAS with artificial intelligence (AI), based on the Industrial Internet of Things (IIoT) technology make further massive strides to both maximise RUL and to prevent asset failure. The benefit: reduction in both unplanned downtime and unexpected replacement of physical assets. 
  • It will not only have data on what broke and how it broke, but also on what had caused the failure, since the data flow registers all asset behaviour. For example, surges in vibration can, through AI, be linked to subsequent gasket and seal failures. Therefore, AI enhances preventative maintenance since then it is not just a case of replacing the seals, but sourcing and eliminating the cause of vibration surges. The benefit: minimising not only asset failures, but also the triggers or causes of such failures. 

In addition, there are further gains on capital expenditure due to better informed Asset Investment Planning for the asset inventory. 

De Gouveia says IBM’s MAS trademark software and algorithms root asset reliance and APM solidly in data science. As a patient’s heartbeat is constantly monitored after open-heart surgery, the continuous flow of data from all instruments assesses deviations from the norm. 

Through machine learning the ongoing interpretation of the digital data flow and assessment against historical operational data and OEM specifications, a single spike (for example, a patient moving in his hospital bed) will be treated as a false negative if all indicators immediately return to accepted parameters. The spike remains recorded and forms part of the performance history of that pump. Subsequent repeat spikes could therefore indicate asset stress and potential failure.
Dashboards automatically portray deviations that can (within predefined parameters) trigger automated remedial actions or highlight and advise the reliability engineer on options for remedial action. The decision-making process in a potentially risky situation is thereby accelerated with AI-informed advice readily available. 

Not using AI in production is akin to budgeting with an abacus 

Turner says to optimise value creation in asset-intensive industries, the investment, application, performance and maintenance of physical assets should be handled as a core strategic business objective and not merely as a support function. 

According to Turner, the problem is that many firms haven’t as yet transformed their operations around data-driven practices and “are stuck in the mode of personnel-driven actions triggered by managerial decisions (Figure 2 describes how the service manager will decide what to do) or by calendar (when is a service due?). Not embracing data-driven decision-making is akin to continue using mechanical calculators, pen and ink to compile budgets and financial statements.” 

“Given the strides in technology, it is strange that, in many asset-intensive industries, the investment in new physical assets, as well as the maintenance of the asset base, are treated as an afterthought.  These are the most fundamental core functions and should be actively aligned with the corporate business strategy and objectives. 

This outmoded approach is still governed by the paradigm of the Model T Ford: service and replace regularly in accordance with the service cards or by weeks, but in case of malfunction, in between…” 

Figure 2 Extract from the 1925 Model T Ford Service Manual 

Just as companies cannot stay on top of their finances by not using spreadsheets and advanced financial systems, ignoring data-driven technologies and asset performance management systems will only entrench an inability to keep up with the pace. 

Turner says the lack of data-layered infrastructure is a common factor at plants that fail to unlock substantial value from both existing data and the technologies offered by the Industrial Internet of Things (IIoT). 

IIoT pilot projects delivered effective streams of digital data. However, problems emerged when scaling up: sometimes due to technical reasons but more often because of a lack of alignment with corporate objectives and insufficient data interpretation capacity due to systems that struggle to handle the data explosion. 

To succeed at this level requires a data-layer strategy capable of handling the data intensity whilst delivering an instant assessment of the performance of assets measured against the performance parameters of the application and the manufacturers of these instruments and industrial infrastructure. Most importantly, the system should evaluate the data against the company’s operational data and context. 

Many of the first cloud-based analytical systems were either not robust enough to integrate properly with traditional databases, or the synchronisation of operation systems presented some obstacles. 

IBM, for decades one of the leaders with its Maximo applications for the industry, provided a basis for Enterprise Asset Management (EAM). With MAS predictive maintenance is enabled to reach “EAM Extended”. 

In a subsequent blog Turner & De Gouveia will discuss the features of MAS: 

EAM Extended: Adding value with IBM’s MAS and Maximo Application Suite is no Rubik Cube: flexibility in a container