Will Humanoids save battery recycling?

Five years ago, we warned that the world was building battery recycling capacity faster than batteries would reach end-of-life. At the time, forecasts and investment theses relied heavily on battery production and sales curves—implicitly assuming that rising demand for batteries would soon translate into rising volumes of battery waste. We now see the consequences of that assumption: pre-processing facilities running below capacity, material recovery projects delayed or paused, and investors reconsidering whether recycling can scale as quickly as promised.

None of this happened because electrification slowed. It happened because batteries stayed in use longer than expected, were redeployed in other applications, or simply moved to other regions. Sales growth did not convert into feedstock. The challenge today is not relevance, but timing—and how capital was deployed ahead of volumes.

This context matters now, because the same patterns are re-emerging in new narratives.

From overcapacity to upstream value capture

A notable and positive development is that many recycling companies are moving toward reuse, repurposing, and services closer to where batteries operate. This shift is often presented as evidence that second-life markets are expanding. In reality, it reflects something more fundamental: recyclers are moving upstream because that is where supply and value can be accessed.

If you can refurbish, resell, or integrate batteries into second-life systems, you can justify paying more for batteries and secure feedstock earlier in their lifecycle. The pivot to reuse is therefore a response to tight feedstock and compressed recycling margins—not evidence of a sudden expansion in reuse volumes.

That distinction matters. Reuse may be the right strategic move, but it does not change the fact that most batteries continue to be more valuable in their original applications for longer than expected. Second-life batteries still need to compete with new alternatives on cost, performance, reliability, testing, and warranty. Growth in downstream markets doesn’t automatically improve their economics.

Why AI and energy storage growth don’t change second-life fundamentals

The latest narrative driving circularity expectations is that the rapid expansion of AI and hyperscale data centers will create unprecedented demand for energy storage—and that this demand will finally make second-life batteries economically attractive.

The first part may be true: grid and behind-the-meter storage could grow faster due to the power and grid constraints of AI clusters. But growth alone does not shift the competitive position of reused batteries. Companies building multi-billion-dollar compute infrastructure are not constrained by battery cost; they are constrained by land, power availability, and grid interconnection. A higher rate of deployment doesn’t mean the market suddenly prefers used packs over new ones.

Faster demand increases market size, not market share.

Second-life batteries still win only if they are cheaper, available, tested, and deployable at scale—not simply because more demand exists.

Humanoids and future applications: new narratives, same logic

A related argument now emerging in presentations and pitch decks is that future categories—like humanoid robots—will become major sources of both battery demand and recycling feedstock.

This repeats the same analytical error seen years ago with electric vehicles: using emerging demand to justify downstream circularity before the installed base exists. The issue with humanoids is not the technology—it is the absence of deployment. There is not yet a proven product-market fit, let alone volumes that could influence recycling economics.

Even if large-scale adoption occurs, batteries in humanoids are small relative to EVs and would not materially shift end-of-life availability for many years. Future demand does not produce present supply.

Narratives move faster than batteries.

Circularity scales differently upstream and downstream

A key misunderstanding in both recycling and reuse discussions is that circularity is assumed to grow when downstream capacity grows. But there are two different dynamics:

• Downstream circularity—reuse and recycling—scales on the availability of end-of-life batteries.

• Total circular value scales on how effectively batteries are managed throughout their lifetime.

Circularity is not just about what happens when batteries die. Diagnostics, warranties, trade flows, controlled redeployment, and planned replacement strategies allow both batteries and their applications to stay in their highest-value roles longer. These activities increase value capture across the lifecycle even when they delay recycling volumes.

This is not a contradiction. It is how circularity works when framed around value, not waste.

Investment decisions must shift from narratives to timing and unit economics

Capital continues to enter the battery industry through thematic lenses: energy transition, decarbonisation, strategic autonomy, AI, robotics. But recycling operations do not generate returns based on themes—they generate returns based on cost and timing.

Capital-intensive recycling facilities only succeed when CAPEX and OPEX align with a revenue ramp fueled by feedstock that arrives soon enough and cheaply enough to generate positive unit economics. If batteries arrive later than expected—or only at prices that compress margin—the runway extends and capital is consumed before utilisation stabilises.

This is not a hypothetical risk; it is what we are witnessing now.

Investors evaluating recycling, reuse, or hybrid models need to look past growth narratives and assess:

• when batteries actually leave service

• where they go next

• whether they retain resale or export value

• how long they remain profitable in first-life

• and who controls access to them when they do retire

Timing—not themes—determines viability.

Why we model the lifecycle, not just the demand

For these reasons, we treat batteries as components embedded within applications, not as independent assets. Batteries live the life of the product they power. They retire when applications retire, not when capacity reaches a theoretical threshold. And many cross borders before doing so.

This is the foundation of the Battery Lifecycle Report inside CES Online: a model built not solely on sales curves, but on usage behaviour, export flows, residual values, repair markets, replacements, and the pathways through which batteries move long before they reach a shredder.

Understanding those pathways is how recyclers improve access to feedstock, how reuse companies create competitive positions, how OEMs preserve value, and how investors avoid funding capacity too early or in the wrong regions.

Circularity will scale. But not as fast as the narratives suggest, and not without deeper alignment to the timing of when batteries truly become available.

Hans Eric Melin

Founder CES Research & Consulting