For embedded hardware teams selling into the EU, the Cyber Resilience Act is the regulation everyone is talking about right now. The Ecodesign for Sustainable Products Regulation is the one they should be talking about next.

The ESPR entered into force in July 2024 and the European Commission published its first Working Plan in April 2025 ,covering the priority products for which delegated acts will be drafted between now and 2030. Energy-related products, including a wide range of consumer and industrial electronics, are in the first wave. The Digital Product Passport, which the regulation introduces, [ESPR-02] will begin rolling out from 2026.

If you are designing connected hardware on a 12–24 month product cycle today, [ESPR-01] the device you tape out in late 2026 will likely be required to ship with a Digital Product Passport. The architectural decisions you make at EVT ,flash budget, modular form factor, BOM traceability, firmware lifecycle ,will determine whether compliance is a configuration change or a hardware revision.

This post is a strategic primer for engineering leaders. It covers what the ESPR actually changes, why modular embedded design is becoming a regulatory advantage rather than just an engineering preference, and the decisions hardware teams should be making now to stay ahead of the timeline.

What the ESPR Actually Mandates

The ESPR replaces the older Ecodesign Directive (2009/125/EC) with a broader, harder-edged framework. Three changes matter for embedded hardware teams.

Mandatory ecodesign requirements for almost all physical products. The older directive applied mainly to energy-using products. The ESPR extends performance and information requirements to almost any physical product placed on the EU market, with delegated acts defining specifics per product group. For electronics, the priority groups include consumer electronics, IT/data centre hardware, and a broad category covering connected and embedded devices.

The Digital Product Passport (DPP). Every product covered by an ESPR delegated act must carry a DPP ,a digitally accessible record of the product’s material composition, repairability information, recyclability data, and lifecycle attributes. The DPP must be machine-readable and accessible by consumers, regulators, repairers, and recyclers throughout the product lifetime. For an embedded device, this means the device itself, or its packaging, must carry an identifier (QR code, NFC tag, RFID, etc.) that resolves to the DPP record.

Performance requirements for durability, repairability, upgradeability, and recyclability. Delegated acts will set minimum performance scores across these dimensions. Products that score below threshold cannot be placed on the EU market. For embedded hardware, this translates into specific architectural requirements: spare-part availability for a defined period, software update support over the declared lifetime, ease of disassembly, recyclability of major components, and recoverability of critical raw materials.

The ESPR also introduces a ban on the destruction of unsold consumer products (textiles and footwear from 2026, [ESPR-03] electronics likely to follow), which has supply chain and inventory implications for hardware companies running consumer SKUs.

Why Modular Embedded Architecture Is Now a Compliance Strategy

For two decades, modular embedded platforms ,PC/104, SMARC, COM Express, Qseven, OSM ,have been a niche engineering choice. Their main appeal was scaling silicon roadmap risk across many product variants and shortening time-to-market for derivative products.

The ESPR rewrites this calculus. A device built around a SMARC module is, by construction, easier to argue as repairable, upgradeable, and recoverable than a device with a soldered-down SoC on a custom carrier. The compute module can be replaced when the silicon ages out. The carrier can be revised when interfaces change. The critical materials in the module are concentrated in a known, separable assembly ,making material recovery measurable.

This does not mean every embedded product should pivot to a modular architecture. The trade-offs remain real:

• [ESPR-04] BOM cost is typically 15–30% higher than a fully integrated design at unit volumes above 10,000
• Mechanical envelope is constrained by the module’s form factor
• Power efficiency at the system level is often worse than an integrated design optimised for the application

What has changed is that the non-engineering arguments ,regulatory defensibility, end-of-life value recovery, support lifecycle simplicity ,are now strong enough to swing the build-versus-buy decision for products that previously would have been integrated.

The teams who will find this easiest are the ones already using modular platforms for time-to-market reasons. We wrote about this dynamic before the ESPR landed, in Accelerating Time-to-Market with Modular Embedded Platforms. What was an efficiency play is becoming a compliance play.

The Six Architectural Decisions That Determine ESPR Readiness

When we evaluate a client’s hardware design against the emerging ESPR delegated acts, six decision points come up consistently. Get them right at EVT and the compliance work is configuration. Get them wrong and the next product revision is a hardware redesign.

1. Form Factor and Disassembly Path

How is the device assembled, and how would a repairer or recycler take it apart? Glued enclosures, soldered batteries, and inseparable assemblies will fail repairability scoring. The bar emerging from sector-specific delegated acts is that critical components ,battery, display, primary PCB, connectivity module ,should be separable by a trained technician without specialised tools, in under a defined time threshold.

For battery-powered devices, the EU Battery Regulation (separate from but adjacent to the ESPR) [ESPR-05] already mandates battery removability from 2027 for most consumer categories. The disassembly question is no longer optional.

2. Spare Parts Strategy

How long will spare parts be available, and at what price? The ESPR is moving toward minimum spare-part availability windows of 5–10 years post-end-of-sale for many electronics categories, with maximum delivery times measured in business days rather than weeks.

For embedded teams, this means the BOM decisions made at EVT now carry a 5–10 year inventory tail. Specifying a single-source component with a 24-month lifecycle is no longer a manufacturing-only concern ,it is a compliance risk.

3. Firmware Update Lifecycle

How long will security and functional firmware updates be provided, and how is that committed to the customer? The ESPR overlaps with the CRA here. The CRA requires security updates over a declared support period. The ESPR will require that period to be meaningful ,likely 5 years minimum for most connected products, with longer windows for industrial and medical-adjacent devices.

This makes the architecture of the OTA update system a 5–10 year commitment, not a 12-month one. The decisions at EVT ,partition strategy, bootloader, signing infrastructure, fleet telemetry ,must hold up for the entire support period.

4. Critical Raw Material Traceability

The DPP will require disclosure of critical raw material content. For embedded devices, this primarily means rare earth elements in magnets and sensors, tantalum in capacitors, cobalt in batteries, and increasingly the conflict minerals tracked under existing regulations (tin, tungsten, gold).

This is a supply chain transparency requirement that starts at BOM commit. Your assembly house knows what they put on your board. The question is whether you can extract that data programmatically into a DPP-compatible record. Most embedded teams cannot today.

5. Recyclability and Material Recovery Design

Can the major assemblies be separated into clean material streams at end-of-life? Multi-material enclosures with embedded electronics fail this. Designs where the PCB can be removed cleanly, the battery isolated, and the housing materials sorted by polymer pass it.

Material recovery is where the European Commission’s Raw Materials Initiative becomes directly relevant: increasing the share of critical raw materials recovered from end-of-life electronics is a stated EU policy goal, and ESPR delegated acts are the implementation lever. Designs that anticipate this are scoring well in early consultations.

6. Energy Efficiency Over the Declared Use Profile

For energy-related products, the ESPR will continue and extend the older Ecodesign Directive’s efficiency thresholds. For embedded devices, this is now interpreted against the declared use profile ,not a single benchmark. A connected sensor declared to operate 99% in sleep mode is measured against that profile, not against peak active power.

This rewards designs with aggressive sleep states, well-tuned wake sources, and accurate power budgets validated under realistic duty cycles. It penalises designs where “we’ll optimise power later” is the EVT philosophy.

Where the Service-Based Business Model Opportunity Sits

The ESPR creates a regulatory tailwind for business models that previously struggled to gain traction in embedded hardware.

Hardware-as-a-Service / leasing. When the manufacturer retains ownership of the device, the disposal pathway is controlled, end-of-life material recovery is realistic, and the firmware lifecycle commitment becomes part of the service rather than a compliance afterthought. The ESPR favours these models structurally.

Take-back programs. Manufacturers that operate take-back programs gain access to end-of-life material flows, which becomes valuable both for compliance reporting and for actual material recovery as critical raw material prices continue to rise.

Refurbishment and second-life. Devices designed for repairability ,modular architecture, replaceable batteries, separable assemblies ,have a credible second-life market. The economics of refurbishment improve when the regulatory cost of disposal rises.

The teams we see executing on these business models well are not the ones who pivoted to them in 2024 ,they are the ones who designed for them at the product architecture stage in 2022 and 2023. The hardware architecture sets the ceiling on what the business model can become.

What to Do in the Next 12 Months

For products in active development today, three steps are practical and high-leverage.

Step 1 ,Scope an ESPR readiness assessment for products currently between PoC and DVT. The cheapest time to make the architectural changes is before tooling. Identify the gaps against the six decision points above. Prioritise the ones that require hardware changes (form factor, spare parts strategy, disassembly path) over the ones that are firmware-changeable (update lifecycle, energy profile tuning).

Step 2 ,Begin DPP infrastructure work. The DPP is data infrastructure as much as it is a regulatory requirement. The product needs a unique identifier resolvable at any point in the lifecycle. The manufacturer needs a database that holds material composition, repairability scoring, firmware version history, repair history, and end-of-life routing data per device. This infrastructure can be built incrementally. It cannot be retrofitted at scale 18 months from regulation taking effect.

Step 3 ,Audit the BOM for critical raw material traceability. Ask your assembly house for material declarations down to the substance level for the priority components. If the data is not available, identify the supplier relationships where it would need to be. This is a supplier conversation that takes 6–12 months to mature.

Where Better Devices Sits

The ESPR is a slower-moving regulation than the CRA ,its impact is felt at the product architecture stage rather than the incident response stage. For embedded engineering consultancies, that means the value is in the design phase: architecting products that are compliant by construction rather than retrofitted.

We work with hardware teams on the design-stage decisions that compound: modular architecture selection, spare-part lifecycle strategy, OTA update infrastructure that will hold up for 5–10 years, BOM choices that support critical raw material traceability, and disassembly path design for repairability scoring.

Our Sustainable by Design piece covered the engineering principles before the ESPR turned them into requirements. This piece is where those principles become a regulatory roadmap.

Frequently Asked Questions

At Better Devices, we work with hardware teams on the architectural decisions that determine whether a product is [ESPR-07] compliant by construction or retrofitted under deadline pressure. Our approach to ESPR readiness starts at PoC and is built into every engagement through to mass production. Related reading: Sustainable by Design, EU CRA Draft Guidance: What Hardware Teams Must Do Before September 2026, and Accelerating Time-to-Market with Modular Embedded Platforms. If ESPR is on your product roadmap and you would value a conversation about where the architectural decisions sit, Book a 30 min consultations with our experts.