Clivet Builds IoT Platform to Monitor Equipment Performance and Minimize Maintenance with OMNYS on AWS

Italy-based Clivet is set to cut maintenance costs by 10 percent using Amazon Web Services (AWS). The company is now ready to build in more cost and operational efficiencies for the residential and commercial customers it serves worldwide with heating, ventilation, and air conditioning (HVAC) equipment. Clivet chose AWS Partner OMNYS to help it build an enterprise-class Internet of Things (IoT) platform on AWS to remotely manage customer installations and to use machine learning (ML) to help predict equipment outages. Using AWS, the company now has a scalable platform capable of ingesting 100 million data items daily.

Clivet, based in Italy, designs and manufactures a wide range of specialist heating, ventilation, and air conditioning (HVAC) products and systems. The residential sector accounts for 45 percent of its business, with the remaining 55 percent of its customers being commercial, such as hospitals, shopping malls, cinemas, and large industrial sites. Its products are fitted with Internet of Things (IoT) devices that can generate hundreds of readings per device every day. These measurements include equipment usage, internal building and ambient temperatures, and gas pressure in air-conditioning units.

The company faced twin challenges. About 35 percent of its customers are based in European countries outside Italy—with almost half of these being in locations where the company does not have an on-the-ground presence—making maintenance issues time consuming to fix. In addition, Clivet’s customers were increasingly turning to the company for equipment maintenance support and requesting help in improving the performance of their HVAC equipment.

The company was using a software-as-a-service (SaaS) solution to track all of the measurements generated by its IoT devices. Because its SaaS system allowed only partial data access, it couldn’t get the complete picture needed to monitor equipment performance. Clivet needed full data access and control to develop better remote management and to train machine learning (ML) models to meet its business goals. The company decided to use modern cloud services and turned to AWS Partner OMNYS, a longstanding IT consultancy that recommended that Clivet build an enterprise IoT platform on AWS. OMNYS estimated that the new serverless platform would need to be flexible enough to ingest 100 million new device measurements daily, with the ability to scale further.

To transfer data between devices in the field and the IoT platform, Clivet uses AWS IoT Core to easily and securely connect devices to the cloud and to authenticate and encrypt all data. OMNYS set up a hot storage system for the last 3 months of IoT data so Clivet can quickly analyze and aggregate trends using Amazon OpenSearch Service, which allows for near-real-time search, monitoring, and analysis of business and operational data. Older data is automatically stored on Amazon Simple Storage Service (Amazon S3), object storage built to retrieve any amount of data from anywhere, so Clivet can easily restore data as required.

All IoT data is then sent to Amazon Kinesis, which the company uses to collect and process near-real-time data streams. To help Clivet predict outages and perform preventative maintenance, OMNYS developed ML models on AWS. For this, it uses Amazon SageMaker, designed to build, train, and deploy ML models for any use case, achieving an accuracy rate of 75–85 percent in the initial test phase. “We’re still working on iterating and improving Clivet’s ML models,” says Davide Pozza, chief technology officer (CTO) at OMNYS. “Using ML models, Clivet has already seen a potential 10 percent reduction in maintenance costs—and many more cost and efficiency benefits are expected.”

Using AWS, Clivet can easily monitor the efficiency of its HVAC units, including for electricity use. It can also easily scale its platform to ingest the ever-increasing volumes of IoT data that its devices generate. “We’ve gathered a lot of data over the last 5 years—we’re now up to 50 billion data items,” says Luca Micheletto, business unit manager of digital solutions at Clivet. “Because AWS allows us to scale, we don’t have to worry about any surge in data growth, because the system just handles it without intervention.”

Cortesy: Clivet

Clivet also wanted to be seen less as an engineering company and more solutions-led. Doing this meant convincing multiple internal engineering stakeholders that this was the right direction for the company. OMNYS was crucial in successfully doing this. “OMNYS has supported us greatly on the journey to create a platform that meets the aims of the business,” says Micheletto. “Having a partner that really understands our business mindset was crucial to convincing our internal teams of this project’s potential.”

Clivet intends to work with OMNYS to further develop its ML models and gain more insights into its IoT data. To do this, OMNYS has designed and built two interfaces to monitor and manage the HVAC systems, including dashboards on AWS that support remote troubleshooting and a smartphone app for end users to review system performance or remotely alter items such as building temperatures. It also plans to benefit from its reliable AWS infrastructure to help it perform over-the-air firmware updates for its IoT devices, further reducing the need for on-site maintenance. “Using AWS means we can better monitor what’s happening in any commercial or residential setting at any given moment,” says Micheletto.

With its new IoT platform, Clivet now has the foundation it needs to establish itself as a solutions-led organization. “Building such a strong and strategic relationship with OMNYS has turned out to be very forward-thinking for us,” says Micheletto. “They understand our needs quickly and are a very proactive partner—we look forward to working with them on the next phase of our development.”