Heavy industry produces tailings, and tailings sit. A ferrochrome plant running for half a century accumulates a stockpile that everyone in the industry knows is there, knows contains recoverable metal, and treats as too difficult to process because the material does not behave like the fresh ore the plant was built to handle. Weathered sludge that has been outdoors for decades has no stable chemistry, no predictable particle behaviour, and no precedent at industrial scale to copy from. The engineering that lasts in this sector is the control system that can correct continuously against a feed nobody has ever industrialised, run well enough that the concentrate clears separate national export specifications. One engineer who built that was recognised in the Best in Business Awards 2025, in high-tech industrial automation.

Yerzhan Temerbayev took the 2025 recognition as a leader in high-tech industrial automation solutions. He has twenty-nine years in industrial engineering and nine years leading Trade-Engineering, the firm he founded. The flagship project was a flotation complex at Donskoy GOK in Khromtau, delivered for JSC TNK Kazchrome, the world’s largest producer of high-carbon ferrochrome, against 14.5 million tons of chromium sludge that had accumulated over more than fifty years of production, with a team of more than fifty specialists and an investment above forty million dollars. It was the first direct chromium-sludge flotation at industrial scale anywhere. The control architecture was built on Siemens hardware, a CPU 1517H-3 PN redundant processor and ET 200SP distributed input and output, running a fault-tolerant SCADA design with real-time monitoring, automated flotation control, integrated safety interlocks and production data feeds. Trade-Engineering has also delivered automation across other sites in the same industrial group, including the Aksu and Aktobe ferroalloy plants and the Kazakhstan Electrolysis Plant, and non-group sites including JSC Pavlodarenergo and the Ekibastuz combined-heat-and-power station. The figures here are as the submission to BIBA 2025 reported them; the programme published its criteria but the individual scores stayed private.

The reason the project resists a confident summary is that the hardest part is the feed. A control system designed for fresh ore can hold a setpoint because the input is consistent. A system fed weathered sludge from a fifty-year pile has to assume nothing is stable and correct continuously, and it has to do that for export-grade product, with a pilot run being no defence. The concentrate is exported to Europe, South Korea, Japan and the United States for high-quality steel, and the grade is set by those buyers and recorded at the customs gateway, where the integrator has no say. The flotation route itself was the unproven part: direct flotation of chromium sludge had never been done at industrial scale, so there was no prior plant to copy a control philosophy from, and the forty-million-dollar investment was committed to a process whose feed behaviour had to be learned in production.

“The material had been accumulating for fifty years,” Temerbayev has said. “The question was not whether the chromium was there — it was whether the chemistry could be made to work at industrial scale on a feed this oxidized, and whether the automation could handle the variability. We answered both. What was unusual about this project was not the idea. It was the combination of process design and engineering execution that made it viable at this scale and at export grade.”

The juror who read that file for whether every figure traced to a signed document was Dmytro Shmatukha. He is a member of the World Business Angels association, sat on the Glonary Awards jury in 2024, and has two peer-reviewed papers in Current Research from 2020 and 2021. He runs three businesses across two sectors and two countries: Angelo, a Ukrainian premium children’s-furniture company he has led as chief executive since March 2016 with roughly half the domestic market and exports to Moldova, Romania, Georgia and Poland, and two US appliance-repair entities, WeFix-Appliance since April 2023 and Swift Appliance Repair since May 2024, the latter growing twenty-five percent between June 2024 and January 2025 on a same-day arrival service promise.

“What struck me about this submission was that every number in it was defensible,” Shmatukha said. “Fourteen and a half million tons is a geological measurement — it comes from decades of accumulation records. The project value is a contracted figure. These aren’t marketing estimates. The engineering team processed something that had been sitting there for fifty years and made it commercially recoverable.”

Shmatukha reads an automation file the way he runs his own businesses, which span two industries on two continents and resolve to one rule. A container of children’s furniture leaving Angelo for Moldova or Poland either meets the receiving retailer’s quality check on arrival or it does not ship again; the same-day arrival promise at Swift either holds on a given job or the dispatch log records that it did not, and the twenty-five percent between June 2024 and January 2025 is a number that dispatch log carries job by job. A forty-million-dollar contract and a 14.5-million-ton stockpile are, to him, the same kind of figure as a furniture container or an appliance call: defensible to the extent every number underneath it traces to a document a named party signed, which in the chromium case means staged acceptance testing and customs records that existed before any award file did. Angelo’s roughly fifty percent domestic share in premium children’s furniture is itself a figure built that way, held since 2016 on the same arrival-spec discipline, which is why he reads an industrial file for its sign-off chain before he looks at its architecture.

The discipline of not claiming a number until the measurement that produced it is the thing holding it up was read by Kirill Timofeev, founder and chief executive of Mad Lead Printer in Belgrade and a managing partner and fractional marketing chief at GFS Group Tradelines, a Los Angeles credit marketplace he has worked since June 2019, where revenue grew roughly elevenfold while the cost of acquiring a customer fell from about 1,700 dollars to about 850. His agency portfolios run around a fifth of client revenue through email lifecycle work, with 45 percent unique-open rates on B2B campaigns and roughly 6 percent cold-outreach response, and his engineering training is in computing technologies for e-commerce.

“When you enter a market without brand recognition, the only thing that creates trust is the work itself,” he said. “Five hundred clients in one year with a 4.96 rating tells you something about how systematically that operation was run from the first day.”

Timofeev reads an engineering claim the way he reads an acquisition file, because the structure is identical: a result is a measurement that held under load, and a result with no instrumentation behind it is unread. The GFS revenue multiple is a different object from the same multiple produced with a rising acquisition cost, and the only way to know which one a file describes is to have measured the cost every month while it ran. A control architecture commissioned through staged acceptance testing carries the same kind of measurement, signed by parties other than the integrator, which is why he reads the chromium file for where the instrumentation lives before he reads what it claims. His own agency numbers are quotable because they were instrumented continuously, and that is the test he applies to a world-first. The architecture is credible to the extent the export grade was being measured by the buyer while the plant ran, with the builder’s later account carrying no weight against the buyer’s record.

The figures here trace outside the builder’s account. A fault-tolerant control system enters production through staged acceptance testing signed by parties other than the integrator. Export grade is recorded at four customs gateways. A commissioning protocol carries the names of the people who signed it. Those records existed before any award file did, which is why a stockpile figure of 14.5 million tons tied to a customs-recorded export grade carries information that a description of innovative architecture does not. A pilot can tolerate variance; product sold into four separate national steel specifications cannot, so the grade is decided outside the plant, by the buyers and recorded at customs, and the automation has to produce that grade continuously from a feed that changes hour to hour.

A circular-economy build of this kind costs the engineer something a greenfield plant avoids. Processing a waste pile that everyone declared too difficult means owning the failure if the feed defeats the control system, and committing to export grade means the buyer’s customs gate decides whether the project worked while the builder’s lab carries no authority over the verdict. The discipline that produces a plant which clears four national specifications is the discipline of designing for a feed you cannot assume, and it shows in the architecture long before any judging starts. The nine years of Trade-Engineering delivery across the Aksu and Aktobe ferroalloy plants, the Kazakhstan Electrolysis Plant, JSC Pavlodarenergo and the Ekibastuz station is the operating base that made the Donskoy GOK risk legible: twenty-nine years in industrial engineering and a team of more than fifty specialists is the difference between attempting a world-first and surviving it. Temerbayev built redundancy and continuous correction because the sludge would not hold still. Shmatukha reads for the signed document because a number without one proves nothing. Timofeev reads for the instrumentation because a claim made before the measurement that holds it up is a hope.

The BIBA 2025 list in these categories also named Abilda Zhunis, recognised for AI integration in logistics and supply chain, and Olga Maximchuk, recognised as Woman Leader of the Year, among its winners. A year from now the citations will have faded and the customs records will not have. Temerbayev’s export grades will still sit at four national gateways, the commissioning protocol will still carry its signatures, and the engineers worth remembering from this cycle are the ones who built the proof into the records other parties keep.