To improve the valuation process of its pre-owned steel, Mlion Corporation partnered with HBLAB to automate the process through an AI-enabled image analysis mobile app.


As it continues to serve as the backbone of many industries in modern society, steel is known as the metal that changed the way we built our world. It allowed us to transform the way we travel, the way we construct our homes and the way we model our cities.

Because of the material’s versatility and strength, estimates suggest the world currently uses more than 1700 million tons of steel per year. Interestingly, 85 percent of this demand is recycled and reused at the end of each life cycle—making steel one of the world’s most sustainable materials. However, issues remain with decarbonising its production. At the moment, the production of steel contributes to around 8 percent of the world’s total carbon emissions. 

Foundation solutions company Mlion Corporation is determined to champion a greater shift towards decarbonization. In an attempt to reduce their company’s carbon footprint and increase efficiency, Mlion Corporation is making efforts to digitalise its processes in trading pre-owned steel. With the help of IPI’s Technology Scouting Programme, Mlion partnered with HBLAB, a software and AI solutions firm headquartered in Hanoi, to automate the valuation process of pre-owned steel through an AI-enabled image analysis mobile application.


A new way to shop for steel

Mlion provides customised products like sheet piles, steel pipes, tie rods and other steel products. Before getting in touch with IPI, the team at Mlion was looking to automate two key processes, corrosion grading and the dimensional measurement of construction steel, in an effort to commercialise a new platform called GoListid. The platform was developed by Mlion to be a one-stop service for the valuation and trading of pre-owned steel.

Pre-owned steel evaluation, which involves measuring the physical dimensions of the steel items and assessing their rust condition, is commonly carried out manually. This process is done by analysing photos provided by the original steel owners and through an on-site inspection of the items. The process is not only tedious, but also heavily reliant on the technical expertise of Mlion’s staff.

We wanted to develop an AI-powered solution to increase efficiency for our team when it comes to analysing and identifying varying stages of rust conditions on steel materials. At the same time, we want to provide transparency to customers when measuring large amounts of material during the production inspection process,” shared GoListid Vice President Brian Wong.


Finding a match

IPI published Mlion’s Tech Need in its marketplace last August 2022 and reached out to numerous AI, machine learning and computer vision solutions-providers that specialise in inspection applications. Respondents to the published problem statement came from both Singapore-based as well as overseas solution providers.

According to Wong, his team was won over by HBLAB’s attention to detail in their project proposal. They were particularly appreciative of the effort HBLAB took to develop a milestone timeline approach to evaluate the AI model’s accuracy.

We were impressed by HBLAB's detailed and contextualised proposal that demonstrated their understanding of our problem statement,” he said. “They also demonstrated technical proficiency by highlighting the methodologies and limitations for solving our specific problem.”

In their proposal, HBLAB also outlined various options including cloud-based and on-device approaches to address Mlion’s needs. Apart from technical solutions, HBLAB presented administrative details like the proposed team structure, communication plan, user acceptance test, warranty information and cloud server maintenance.


Strengthening the bond                                                                                           

“IPI has been very helpful with sourcing and identifying potential technology solution providers for us and sharing their own professional judgement during the vendor evaluation process,” said Wong. HBLAB, on the other hand, is grateful for the opportunity to collaborate with a leading Singapore firm in Mlion Corporation.

Both companies are focused on the next phase of the collaboration; to continue to develop and improve the accuracy of the AI model, further highlighting the potential of IPI’s Technology Scouting Programme in forming long-term business partnerships.