Expertise

Maint.ai

End-to-end sensor-to-action analytical models to dynamically optimise predictive strategies, plant reliability and maintenance response.

Intelligent asset management

Predictive maintenance – six steps for effective implementation

Download article

Deploy smart strategies and models to reduce planned maintenance by 20-40%, unplanned production trips and subsequent flaring by up to 30%, and achieve results in as fast as 12 weeks.

New and existing assets are being challenged to dramatically reduce operating costs (~40%), site staffing (~60-80%), emissions (0), and safety incidents (0) while improving production throughput and availability (~10%). At the same time, ~30% of sanctioned value is lost in the industry on average due to production shortfall and unforeseen costs.

Our intelligent asset optimisation specialists work closely with customers to understand and address the main causes of these issues through operation; deploying smart, ‘low-touch’ strategies, IoT sensors and prescriptive equipment models to accelerate the transition to remote operations and automated maintenance management.

Start optimising

Human intelligence + machine learning + advanced analysis = better business logic for your assets

Improve plant reliability

Target critical systems, equipment and failure modes through prescriptive recommendation models, powered by our industry database

Reduce emissions

Predict and optimise power generation and gas-fired systems, trip-related flaring and methane leaks

Elevate safety

Apply smart strategies and monitoring technologies to remotely diagnose and predict failures while reducing operational cost

The challenge

Many companies are grappling with supply chain disruptions and aggressive operational performance targets. An end-to-end perspective of maintenance operating and data models can vastly improve supply chain processes and operational performance, but optimisation initiatives often encounter these typical problem areas:

Strategies

Strategies are generically derived with heavy reliance on equipment vendors, common sensor and data gaps, without utilising in-field equipment data and not optimised based on business targets and performance.

Prognostics

Prognostics are often conducted manually, critical failures are not detected and eliminated, and predictive maintenance approaches can lead to alarm flooding and false positives, reducing buy-in from the engineering technical authorities.

Execution

Execution challenges are exacerbated further with growing backlogs through queuing and risk prioritisation inefficiencies, maintenance compliance issues and ineffective maintenance resulting in defects on install.

With the increasing rate of learning and technical innovation pushing best practice and increased availability of large equipment data sets across the industry, traditional approaches to managing equipment maintenance come with high opportunity costs.

However, to unlock the value in each of these areas, your business, digital and engineering technical authorities will need to be comfortable that any risks associated with implementing AI are effectively managed.

Wood’s Maint.ai models apply the latest in industry global best practice, lessons learned and value proofs, combining engineering first principles fundamentals and AI models to gain the confidence of your engineering function, to rapidly optimise your maintenance operating model. Our approach allows you to make decisions today with your current data while simultaneously mapping an improvement roadmap with you to achieve your performance targets.

Maint.ai pairs AI with our codified domain expertise, built on your data and embedded in your processes to help your teams manage equipment maintenance and operations end-to-end

1

Apply globally vetted, best-in-class models, benchmarks and hypothesis-driven problem solving that we test on your data

2

Integrate and augment your maintenance processes, aligning with your meetings and key decision requirements

3

Work together with your key engineering technical authorities from day 1, finding the critical bottlenecks and risks preventing a decision

4

Fill your data gaps with our maintenance quality enrichment models, testing and triangulating multiple sources to search for the truth

5

Use an agile implementation framework designed to accelerate value realisation from operations transformation initiatives

For some of our clients, the pace of transformation can be exhausting – especially when fire-fighting reactive issues. To accommodate this, our models are deployed client-side and embedded in your organisation, along with training and performance measurement systems to track their effectiveness.

This puts you in the driver's seat to be able to complete the remaining changes at your pace. Alternatively, our domain experts are happy to help eliminate issues for good, as you focus on improving the overall system.

Maint.ai key features:

Rapid evaluation on large datasets and benchmarks against our global database and models
‘Low-touch,’ remote maintenance strategies defined based on failure mode to symptoms mappings
Explainable equipment thermodynamics & dynamics physics models
Sensor prognostics models
High-performance anomaly detection and remaining useful life ML models (>98% recall, >80% specificity)
Data quality enrichment models (CMMS, production delay, time-series data)
Global equipment class MTBM, MTBF, MTTR statistics

Mixed failure mode reliability growth population forecasts and modelling
Failure histories and case/effects of individual equipment
Historical equipment and strategy compliance
Temporal maintenance execution assessments
Failure mechanisms and maintenance types
Standards and regulations
Equipment criticality and redundancy Identification of existing and upcoming low-risk work that can be postponed in lieu of higher risk maintenance
Identification of high-risk, overdue work for prioritisation

What you get

An optimised maintenance portfolio in 8-16 weeks, resulting in a 20-40% reduction in maintenance cost and 10% reduction in unplanned failures.

A clear path to smart maintenance – from the plant business case through to sensor and IoT selection, through to continuous optimisation
Increased accuracy and recall in condition assessments and predictions
Optimised ‘low-touch’, smart maintenance strategy outcomes
Increased critical asset availability
Reduced production losses and trips, as well as trip related emissions
Decreased maintenance cost without production loss
Reduced POB requirements and a workforce focussed on reliability optimisation
Real-time updates of strategy, equipment maintenance and backlog improvement suggestions

See if Maint.ai works for you

Click the button below, and one of our consultants will get back to you shortly to arrange a consultation or demo.

Get started

Wood has been optimising equipment maintenance and reliability of asset-intensive industries for 25+ years.

We have a rich history of innovation and first-principles understanding, having pioneered the field of equipment condition monitoring, working closely with clients and our partners in building sensing hardware and predictive maintenance software to troubleshoot and resolve critical equipment problems.

A recent example has seen our teams co-design our client's wireless vibration sensors, capable of high integration, accurate measurement, and data acquisition rates of 1-15 mins and co-deployed these sensors to 20 of 2,000 fin fans at one of their LNG sites. Our contribution accelerated development and deployments by ~2 years.

Based on extensive client knowledge and 10+years of operational data, we created and deployed explainable artificial intelligence models to production within four months.

We created a wireless sensor strategy assessment framework, capable of rapidly qualifying market sensors and building and developing an integrated architecture.

We have convinced our client's chief engineers to trust our transparent sensing analytics approach and architecture across all seven existing and new assets.

Human + machine: combining domain expertise with data-driven models

Technology-enabled, independent engineers: combining domain expertise with data-driven models
Global leaders: in our disciplines, based on decades of experience
Superior in our capability offering: full complement of in-house expert knowledge and experience
Problem solvers: combination of desktop simulation and ﬁeld experts for optimum solution development
First-class in our service: in-house software tools, resulting in improved accuracy, faster service and lower costs
Thought leaders: based on our strong commitment to research, innovation and knowledge sharing in the industry
Consistent in our delivery: standard processes and quality control for inspections, simulations and risk evaluation

Innovating

Data and analytics trends indicate that industry is increasingly using data fabric approaches to help address complexity and scale their data assets (Gartner, 2021). Data-driven models are only scratching the surface of what is possible.

Backed by Wood and innovation funding, our team is always pushing to find better, faster and more robust solutions. With the ability to build specialised teams from over 40,000 experts in 60 countries, our solutions have no boundaries. See if Maint.ai works for you