To reduce costs and increase production levels for asset owners, P4A saw an opportunity to harness Internet of Things (IoT) and analytics solutions to enable predictive maintenance and to boost asset performance.
In an eight-week engagement, P4A worked with the IBM Garage to build a minimum viable product (MVP) that monitors wind turbines and provides real-time alerts to technicians via a virtual assistant.
Winsbusiness for P4A by helping wind energy producers boost asset performance, increase efficiency and reduce costs
8-weekdevelopment time for MVP using pre-built components and expertise from IBM
ReadiesP4A for growth with highly scalable solution architecture and cloud resources
Business challenge story
Sparking an idea
By 2020, the European Union (EU) is aiming to boost the proportion of energy generated from renewable sources to at least 20 percent. In 2016, wind overtook coal as the second largest form of power generation capacity in the EU, and it’s fast catching up to the leader: gas.
As demand for wind energy grows, producers have the chance to make huge gains—if they can tackle the issues that limit production levels. Apart from weather conditions, which wind energy producers cannot control, the other main factors that affect output are asset performance and availability.
Until now, most wind turbine owners have had little to no insight into the condition of their machines. Even though their end-of-design lifetime is usually close to 20 years, original equipment manufacturers (OEMs) for wind turbine components will typically guarantee operation throughout a 12 to 15-year long-term service agreement (LTSA). Once the LTSA comes to an end, wind energy companies struggle to find insurance for components. At the same time, owners lack visibility of how well assets are performing, meaning they could be operating at well below maximum throughput for years at a time.
Performance for Assets (P4A) saw an opportunity to extract much greater value from wind turbine assets. Laurent Rakoto, Data Analyst at the company, explains: “Effective predictive maintenance could significantly help to extend the life and value of assets. But while lots of energy companies have monitoring solutions that generate huge amounts of data, most of it goes unused. Some include alerts or warning systems, but don’t give field technicians enough information about what actions they should take next. This means they miss the opportunity to significantly raise their output by making sure that every component works to its best possible potential.”
In response, P4A created Wintell, an advanced monitoring system that combines sophisticated analytics with field experts’ process knowledge in preventive, corrective and predictive maintenance, condition monitoring, testing inspection and certification and data mining to provide actionable intelligence. P4A chose to focus on wind farms as an initial use case.
In developing the concept, the P4A team realized that it needed support from technology experts and a flexible, scalable cloud platform to bring Wintell to market successfully. Rakoto adds: “To launch Wintell commercially we needed to extend it so that it could better accommodate big data and machine learning workloads. We also wanted to add new functionality, so we looked for partners that could help us create a marketable solution with leading-edge capabilities.”
Firing up innovation
P4A chose to work with the IBM Garage in Nice to bring its cloud-based advanced monitoring system for wind farms to life. The worldwide network of IBM Garage locations is designed to make IBM Cloud industry knowledge and higher value technologies accessible to enterprises globally. P4A engaged I-Pulses, experts in business intelligence and analytics solutions, for help with the project.
“We learned about the IBM Garage during a visit to the IBM Global Industry Solution Center in Paris,” recalls Rakoto. “P4A couldn’t pass up the opportunity to work with a center of excellence and get experience with the latest technology. I-Pulses brought additional expertise to help us get the data science piece of the solution just right.”
To kick off the project, P4A and I-Pulses participated in a three-day Design Thinking workshop to refine the Wintell concept. During the session, the team defined the target users, their precise needs and how the solution could most help them in practice. Rakoto recalls: “The IBM consultants asked us lots of questions that made us think about aspects of the solution that we hadn’t considered before. The emphasis was very much on how we could add more value in the market, which helped us focus and get to our desired end-result sooner.”
Next, the Garage team employed agile development and continuous delivery techniques to help P4A and I-Pulses create an MVP in an IBM Cloud Foundry environment.
The solution collects and processes data from wind turbine sensors and combines it with weather forecast information using IBM Watson IoT™ Platform solutions hosted in the IBM Cloud. Data is stored in IBM Informix® on Cloud, a fast, scalable database, delivered as an automated infrastructure-as-a-service solution.
P4A data scientists worked with IBM and I-Pulses to build hybrid machine learning models that detect and diagnose issues with wind turbine components, and provide an accurate prognosis to indicate how long each asset will perform well. With the IBM Watson® Machine Learning service, the company can deploy these models into production at scale, taking advantage of automated, collaborative workflows to streamline development.
With support from the IBM Garage technical experts, P4A and I-Pulses created a user interface that allows field technicians and managers to browse analytics results and visualize wind turbine data. It also incorporates a virtual assistant called Nestor using IBM Watson Assistant, which enables users to ask questions about recommended next actions, the status of different components and prioritization of jobs according to their impact on cost, output and availability.
“IBM brought in the right experts at the right time to help the project go smoothly,” says Rakoto. “We were unsure about how much we could achieve in the tight timeframes we set ourselves, but IBM helped the team to exceed expectations. They delivered some functionality we didn’t think could be built in time and were very responsive whenever we asked for help.”
Taking the wind farm sector by storm
Working with IBM and I-Pulses, P4A created an MVP for Nestor in just eight weeks. This was in time for the Euromaintenance 4.0 conference, an event focused on disruptive technologies in asset-intensive industries. It was an ideal opportunity for the company to showcase the exciting new capabilities of its app to its target audience.
“Getting Nestor ready for Euromaintenance 4.0 was a big priority for us, and it wouldn’t have been possible without the support of IBM and I-Pulses,” comments Rakoto. “Choosing IBM Garage allowed us to take advantage of pre-built technology in the IBM Cloud and specialist skills from a global player.”
Equipped with Nestor, asset owners will gain the insights they need to increase output while reducing costs, benefiting their bottom line. The solution could also give insurers the confidence to provide cover for windfarm assets once the service agreements with OEMs have lapsed, by lowering the risk of unforeseen malfunction throughout a component’s lifetime.
“Even small changes to maintenance programs or machine settings can have a big impact on costs and revenues,” explains Rakoto. “Our solution provides real-time alerts so that asset owners can address wind turbine under-performance and unavailability very efficiently. This helps them to get the most out of their equipment and maintenance teams. It also makes it less of a risk for insurance companies to cover windfarm assets that are more than 12 to 15 years’ old, but still a few years from end of design lifetime. As a result, asset owners can choose to delay the replacement of assets, or to decrease costs without lowering production levels.”
P4A is gearing up to launch Nestor commercially. With IBM Cloud resources, the company is confident that it can scale up quickly to meet new demand, enabling it to move fast on opportunities.
Rakoto concludes: “Nestor got a great reception at Euromaintenance, and we’re seeing interest from assets owners. We’re targeting 5 percent of EU onshore wind turbines, which could have their lifetimes extended and their performance boosted. Thanks to IBM Garage, we now have a simple, scalable platform based on IBM Cloud solutions that we can take to market to start making waves in the energy sector.”
About Performance for Assets
Performance for Assets (P4A) serves the energy and industry sectors with actionable insights that enable companies to optimize production, increase uptime and save energy. Established in 2017, P4A is spin-out company based in Belgium, built on a collaboration between its shareholders Maintenance Partners, Vincotte, Icare and SRIW. P4A’s partners offer expertise in the fields of maintenance (preventive, corrective and predictive), reliability (condition monitoring), TIC (Testing Inspection and certification) and data mining.
Take the next step
Learn more about IBM Cloud solutions by contacting your IBM representative or IBM Business Partner, or visiting the following website: ibm.com/cloud
Learn more about IBM Watson Data and AI solutions by contacting your IBM representative or IBM Business Partner, or visiting the following website: ibm.com/watson/products-services
IBM is an established leader in the Internet of Things with more than 6,000 client engagements in 170 countries, a growing ecosystem of over 1,400 partners and more 750 IoT patents which together help to draw actionable insight from billions of connected devices, sensors and systems around the world. For more information on IBM Watson IoT, please visit ibm.com/iot.
I-Pulses has more than 50 years of cumulative experience in business intelligence (BI) and data management, and was established to provide personalized support for clients’ BI and analytics needs. For more information about I-Pulses, please visit i-pulses.be