Unlocking the power of manufacturing data with CloEE

Over the last decade, Industry 4.0 promised to bring data-driven insights to the manufacturing sector, reducing inefficiency, supply hold-ups and logistical bottlenecks, deploying predictive maintenance to reduce machinery downtime, overall saving time and money. 

However, factory digitisation presents many challenges that hinder its seamless implementation. Integrating diverse systems, devices, and technologies, especially bridging the gap between IT and OT systems, poses significant complexity. Substantial initial investments and the difficulty in justifying costs and demonstrating ROI have further complicated the adoption process. 

The process is still relatively nascent, with 64 per cent of manufacturers still in the first stages of digital transformation.

In response, Finnish startup CloEE is an AI digital advisor for discrete manufacturing that boosts sustainability, improves overall equipment efficiency (OEE), and drives annual revenue growth of up to $1 million. 

I spoke to Julia Sabitova, co-founder, COO, and Chief AI Advisor Nick Gushchin to learn more. 

CloEE simplifies data-driven manufacturing insights

According to Sabitova: 

"Manufacturers have a problem with data.  Either they lack data, their data is dirty, or they don't know what to do with it.

Further, there's the challenge of implementing data analysis. 

"Most manufacturing companies have over 20 operators who produce every day and — often night as well, and they just don't have time to train people to use different platforms." 

Discussions with a competitor revealed that workers needed a month's training or more to utilise their software effectively. 

In response, CloEE brings highly accessible digitisation to mid-size manufacturing companies. Companies can start with one facility and then scale it to their other operations. 

Leveraging generative AI, CloEE delivers continuous improvement using real-time data from manufacturing equipment, MES, and ERP systems. 

It reduces energy consumption by 30 per cent and emergency stops by 95 per cent, while providing quick, cost-free project implementation to save manufacturers time and money.

With CloEE, workers don't require deep domain expertise in data analytics, AI, advanced automation or advanced upskilling.

"We realised that everyone knew about data collection systems, but no one used them properly. But we could build a simple-to-use, easy-to-understand and deploy plug-and-play platform with no charge for implementation that could deliver results and ROI in less than three months from installation."

Bringing value to mid-size factories

CloEE's focus on mid-sized factories is strategic. There are over 200,000 manufacturing companies in Europe, the US and APAC, many smaller in size. 

According to Sabitova, the benefits of Industry 4.0 have yet to pass them by. But many still operate in a brownfield-greenfield environment with only a portion of machinery digitised. 

"Manufacturing companies don't have one type of equipment. They have a zoo of equipment.

They want to drive through the implementation of ERP or MES systems. Many of them use SCADA systems, so they have some level of digitalisation. But it doesn't influence the process system significantly."

And in turn, CloEE's tech has gained fast traction in less than two years of operation. 

It offers a number of distinct advantages over its competitors. Workers can not only use it remotely on their phones, but the software speaks their language, meaning workers aren't waiting months for translations.  

Further, CloEE's AI digital advisor can be accessed remotely and even by mobile. This is a huge boost for factories where the IT/OT divide persists. Factory floor data is often relished to the first-floor offices and only accessible via desktop or laptop, meaning machinery workers cannot access the data that benefits them most. 

 "Our platform supports over 100 protocols, and our team has successfully connected over 25000 devices globally."

CloEE's leadership team brings a wealth of experience. Oleksandr Zadorozhnyi, CEO, brings over a decade of global sales experience in Industry 4.0, IoT, and SaaS solutions. Julia Sabitova, COO and co-founder, leverages her extensive B2B marketing expertise to drive CloEE's growth.

Nick Guschin, Chief AI Advisor, has a proven track record in leading AI projects across diverse sectors. Slava Mitin, CTO, brings over a decade of industry experience, having successfully implemented Industry 4.0, IoT, and AI solutions in large-scale manufacturing environments.

Robert Klöpsch, Chief AI Officer, has 10 years in data science & analytics, optimising processes in industry and finance using AI and ML, with a proven track record with Fortune 500 in AI / ML deployment. 

Driving a greener manufacturing industry 

CloEE contributes to sustainability by leveraging existing machinery and equipment.

"We are ready to work with any type of data from equipment or machines."

Further, by leveraging data from equipment and energy consumption, the system enables enterprises to decrease equipment electricity consumption through efficient utilisation while also decreasing scrap, making production more environmentally friendly by using fewer resources and reducing production's CO2 footprint.

Additionally, analysing equipment data provides valuable insights into machine performance, usage patterns, and potential issues like product quality problems. This granular level of data empowers businesses to pinpoint the root causes of issues by identifying specific operators, modes, and timeframes.

"We believe that a comprehensive, 100 per cent equipment data collection system can lead to a significant efficiency boost of 20-30 per cent within just three months."

Real-world perspective of AI costs and challenges

I've been hearing about AIoT and AIIoT for a long time, but few people are willing to share the practicalities of AI in industrial settings, especially when it comes to cost and risk.

According to Gushchin, CloEE initially explored training its own LLM but realised the significant cost, exceeding $2 million. 

"Given the challenges faced by companies like Aleph Alpha, which recently pivoted from their initial sovereign GPT model to Ai support due to high costs and competition, we believe a more sustainable approach is necessary. 

Further, we see the AI market as fragmented, with various players specialising in different areas. 

There will be models, providers and users. Due to the high demand for this technology and the significant investment required, we won't compete with OpenAI or similar companies."

CloEE currently utilises Azure OpenAI service models. "We can seamlessly shift between models within days, or even hours, as we have pre-installed plugins and protocols to facilitate this transition."

I was curious about the security around training LLMs using industrial data.

Manufacturing is currently the most targeted industry for cyberattacks, representing 20 per cent of all cyber extortion campaigns in 2023. An attack that downs operations can result in financial costs of millions per day not to mention reputational damage. 

Gushchin explains that the company uses less sensitive data to train its models. 

Further, "Azure provides us with a robust and secure cloud environment for customer data. Our LLM-agnostic solution can be deployed both in the cloud and on-premises, catering to diverse customer preferences and ensuring data security."

This year, CloEE participated in European SkyDeck Berkeley, an international accelerator program for European-based startups. 

Sabitova praised the program, recalling an influx of 150 emails into her inbox — many via word of mouth: 

"What was even more interesting is that when I replied to the emails, people answered, many immediately. I was not waiting to get a response. People really understand the need we are addressing."

CloEE has gained traction globally, including in Finland, Greece, Italy, Switzerland, Malaysia, and India. 

With 10 million industrial facilities housing approximately 1 billion machines, a vast reservoir of untapped potential exists. The company has seven paid customers, including Hyundai, with 200 connected machines. It aims to reach 90 potential customers ready to deploy its solution and gain a competitive edge next year.

Lead image: Freepik.