Instant Chemistry,
Zero Furnace Idle
Turning a slow, manual spectrometer workflow into a real-time WhatsApp & web experience for foundry teams.
The non-homogenous nature and high impurity content of raw materials in metal manufacturing present a significant hurdle. To meet product specifications, materials must undergo extensive spectrometer testing and precise addition/dilution — a lengthy procedure that drives up production costs as furnaces continuously draw power to sustain the required high temperatures while on hold.
The design process
An eight-step process carried the project from raw field research to a shipped, adopted product.
Understand the existing system
Based on our research, we found that just like a chef tastes his food to perfect the seasoning, foundries analyse molten metal after melting. A bath sample is taken, solidified, and tested with a spectrometer. By comparing the spectro result to the desired final chemistry, adjustments are made through an addition/dilution process to achieve the perfect final product.
Define the system
After our research concluded, we brought together all the data collected by the team. While specific procedures vary across foundries, there are significant commonalities in their overall processes. Below is a general overview of the typical foundry process.
Identify opportunities
The current process of taking a bath sample, sending it for spectrometer analysis, awaiting the test report, relaying it to the melting supervisor for addition/dilution calculations, and then proceeding with the adjustments, is time-consuming. This extended wait keeps the furnace on hold, significantly increasing production costs and leading to a loss of energy, time, and overall productivity.
Ideate for solution
After analysing all our data and conducting brainstorming sessions, we developed an approach to reduce furnace holding time — leading to decreased costs and increased profits. Our strategy: build an MVP first, hit Product-Market Fit by testing with 5 customers, then refine for final release.
Connect the machine with IoT and gather real-time data on our server.
Analyse the data and use AI & ML to calculate the addition/dilution suggestion in real-time.
Showcase the result to the users over WhatsApp.
The web app stores a complete history of data, letting users review past heats.
Create designs & develop
Once we had ideated on a MVP product, we went forward to design the initial interface of the web app, followed by its development for testing and launch.
Usability testing & launch
Once the MVP of the product was ready, we conducted usability tests with actual users to assess product feasibility, and fixed any issues found during the process. Once satisfied with the product's usage parameters, we launched the MVP in beta to validate our hypothesis through market adoption.
Product updates
Upon launching the MVP to the market, our sales and onboarding teams' dedicated efforts successfully secured our initial customers. Reaching the five-customer mark validated our MVP and confirmed Product-Market Fit (PMF). During this crucial validation, we found the following issues.
Introduced a display showcasing live readings and suggestions at the shop floor.
Introduced a dashboard for senior stakeholders with filters to draw insights.
Impact showcase
Since inception, monthly total users have climbed steadily as more foundries adopted SpectroPro.
“With MetalCloud's SpectroPro module, there's no more furnace idle time. We get instant results, adjust immediately, and save energy and costs every day.”
“Melting supervisors can act immediately with MetalCloud's SpectroPro module, without follow-ups or last-minute rush. It's simplified our workflow.”
Every minute a furnace idles on a spectrometer result is a minute of energy and margin lost. Closing that loop in real time turned a manual, multi-step wait into an instant decision.
Shipping a lean MVP, validating with five real customers, then building the shop-floor and dashboard views the data itself demanded — not the other way around.
