In 1976, in Nyunge village near Bukoba, on the western shore of Lake Victoria in Tanzania, a group of Haya elders built a furnace that none of them had fired since they were young. They had last smelted iron in their youth, 50 to 60 years earlier. Now, working with the archaeologist Peter Schmidt and the engineer Donald Avery, both of Brown University in the United States, they did it again.
They raised a cone-shaped shaft from old slag and mud dug out of a termite mound. They burned swamp grass in the furnace bowl until it was packed with charred reeds. Eight drum bellows covered in goatskin pushed air through clay pipes, called tuyères, that reached inside the furnace, so the air was already hot when it met the fire. The bellows ran for up to eight hours. What came out was a bloom of carbon steel.
What the archaeologists found
Schmidt and Avery reported the work in Science in September 1978. They measured the air inside the pipes at around 600°C or more. In many runs the combustion zone was hot enough to melt a platinum-rhodium thermocouple, which melts at about 1,820°C. Then they dug. At an Early Iron Age site near the lakeshore, known as KM2, they found evidence of a similar furnace form and the same preheating. They concluded that people in the area had practised this technology 1,500 to 2,000 years ago.
Not every scholar accepted every claim. In the years that followed, metallurgists such as J. E. Rehder and David Killick questioned how much the preheating really mattered and whether "advanced" was the right word. That debate is healthy. What it does not touch is the simpler point: people on the shores of Lake Victoria were smelting iron and making steel many centuries before anyone arrived to "introduce" industry to them.
Why the fires went out
So why had the elders stopped? The 1978 paper records their own answer. The smelters blamed the availability of cheap, imported iron tools and spring steel, and the better money to be made from growing coffee.
The skill did not die because it failed. It died because it stopped paying. Imported goods arrived cheaper than local ones, a cash crop paid better than a craft, and within a lifetime the knowledge survived only in the memories of old men.
The same story, a different ore
That pattern has a modern version, and it runs on cobalt, the metal inside many of the batteries that power phones and electric cars.
According to the US Geological Survey, the Democratic Republic of the Congo produced an estimated 75% of the world's mined cobalt in 2024. Yet China refined 79% of the world's cobalt that year, according to the Cobalt Institute's 2024 market report. The USGS notes that most of China's refinery output came from partially refined cobalt imported from the DRC and Indonesia. And according to the Oxford Institute for Energy Studies, most of those intermediates travel to China through South Africa, Mozambique or Tanzania. Some of the cobalt leaves through the same country where the Haya once made steel.
Cobalt is not an exception. The International Energy Agency finds that China is the leading refiner for 19 of the 20 energy-related strategic minerals it analysed, with an average market share of around 70%. It expects China to still supply more than 60% of refined cobalt in 2035. We dig. Others add the value, and with it much of the margin, the jobs and the know-how.
Why now
The DRC has begun to push back. In February 2025 it temporarily banned cobalt exports, and in October it replaced the ban with quotas of up to 96,600 tonnes a year for 2026 and 2027. Whatever one thinks of the policy, it shows that producing countries no longer want to be quarries. The harder question is whether we are ready to build what must come next: refineries, battery-material plants, and the thousands of smaller factories that turn raw material into finished goods.
That question is not only about cobalt. It applies to the coffee, tea and hides that too often leave the continent before most of their value has been added.
Lighting the furnace again
The Haya elders left us something more useful than nostalgia. Knowledge that goes unused fades, but it does not vanish overnight. Given a reason and some support, men who had not smelted in half a century ran six smelts in a single year and produced steel.
Value addition works the same way. It is not a slogan. It is thousands of decisions by factory owners, engineers, financiers and policymakers to process one more step at home. It needs reliable power, patient capital, skilled people, and trade rules that let a processor in Kisumu sell to a buyer in Kinshasa. It also needs factories that know their own numbers: what each batch really costs, where the waste goes, and which line is losing money.
Every African plant that processes what we grow and mine is that furnace being lit again. So the question the elders answered in 1976 is still ours to answer: what are we still importing that our grandparents knew how to make?