From Heat to Data: How Energy Harvesting Chips Are Powering the Internet of Things
- jenniferg17
- Nov 17
- 3 min read
Key Takeaways
Why energy harvesting is not a “free battery”
Energy sources and their design trade-offs
System-level architectures for self-powered IoT
Reliability and lifetime factors
Commercial proof of scale
Semiconductor market outlook
Why Energy Harvesting Is Not a “Free Battery”
It’s tempting to imagine energy harvesting as a limitless alternative to batteries, but the engineering reality is far more nuanced. Batteries store predictable energy; harvesters rely on the environment. Most ambient sources deliver typically between tens of microwatts to a few milliwatts, meaning designers must build systems that survive within tight, fluctuating power budgets.
For example, a pipeline thermoelectric generator (TEG) may deliver tens of milliwatts only if the temperature gradient is steady. A piezoelectric harvester performs well on a vibrating motor, yet almost nothing when the frequency shifts. Indoor photovoltaics sustain sensors in well-lit spaces but struggle in shaded corners. RF harvesting can power small tags near routers, but power density drops sharply with distance.
The lesson: energy harvesting is about resilience, not abundance. Every subsystem, from the PMIC to the MCU and communication stack, must be designed to operate reliably under energy scarcity.
Energy Sources and Their Design Trade-Offs
Across these sources, steady environments favor PV and TEGs, while dynamic systems favor mechanical or RF harvesters. Hybrid architectures increasingly combine multiple sources for redundancy.
System-Level Architectures for Self-Powered IoT
Harvesting is only half the challenge, the other half lies in system architecture.
Ultra-Low-Power Microcontrollers
MCUs such as STMicroelectronics STM32U5 and Renesas RE01 operate at tens of µA/MHz and sleep in the nanoamp range. These figures are not incremental—they define whether a node runs continuously on harvested power or shuts down unpredictably.
Dedicated Energy-Harvesting PMICs
Chips like TI BQ25570 and ADI LTC3331 boost millivolt-level inputs, manage supercapacitor or lithium-cell storage, and balance charge/discharge cycles. Multi-source PMICs even blend light, vibration, and RF energy to stabilize supply in fluctuating conditions.
Adaptive Firmware
Fixed duty cycles don’t work when the power supply changes minute by minute. Firmware must adapt dynamically: event-driven wakeups, variable sensing intervals, and local data compression allow systems to operate autonomously for years.For example, instead of transmitting every temperature sample, a node can detect and send only anomalies, cutting radio energy use by orders of magnitude.
This hardware-software co-design defines the new frontier of IoT engineering.
Reliability and Lifetime
A critical but often overlooked question: Do harvesters outlast batteries?
PV cells: <1 % annual degradation under indoor light.
Piezo harvesters: mechanical fatigue is possible after billions of cycles.
TEGs: prone to solder fatigue from repeated thermal cycling.
Supercapacitors: typically 500 k–1 M cycles depending on ESR and temperature but notable leakage currents.
Long-term reliability depends more on material stability and storage design than on conversion efficiency. Engineers increasingly design for degradation—derating components, adding redundancy, and adapting firmware as energy conditions evolve.
Commercial Proof of Scale
Energy harvesting has matured beyond academic demonstrations. Proven deployments include:
EnOcean – PV-powered light switches and sensors in over one million smart buildings.
Everactive – Vibration-powered monitoring nodes operating continuously in industrial refineries with ABB.
Wiliot – RF-powered Bluetooth “sticker tags” used in retail and logistics for asset tracking.
These examples confirm that energy-harvested systems can scale commercially while maintaining reliability.
Market and Semiconductor Outlook
The global thermoelectric generator market is projected to grow from USD 800 million (2023) to USD 1.6 billion (2030). Combined with PV, RF, and piezo solutions, energy harvesting is poised to become a cornerstone of low-maintenance IoT infrastructure—from smart buildings and wearables to environmental monitoring.
Semiconductor manufacturers are aligning rapidly:
STMicroelectronics & Renesas – MCUs with sub-µA standby currents.
e-peas, TI, ADI – Dedicated EH PMIC portfolios.
Silicon Labs – Radios optimized for harvested-energy operation.
Next-generation designs will emphasize co-packaged PMIC-MCU systems, foundry-level ultra-low-leakage processes, and energy-aware simulation tools, reducing both BOM cost and design risk.
Conclusion
Energy harvesting has evolved from research curiosity to a practical design paradigm enabling IoT nodes that operate for extended periods on ambient energy. By integrating optimized harvesters, multi-source PMICs, ultra-low-power MCUs, and adaptive firmware, engineers can build devices that deliver reliability within microwatt budgets.
At McKinsey Electronics, we view energy harvesting as a key enabler for next-generation IoT across industrial, wearable, and smart-infrastructure markets.Through our semiconductor distribution and circuit-design advisory network, we support engineers across the Middle East, Türkiye, and Africa in selecting the right silicon, from efficient PMICs to ultra-low-power MCUs, to transform ambient energy into sustainable intelligence.
Sources
B. Li et al., A Thermoelectric Generation System for IoT Using Waste Heat from Pipelines, arXiv:2203.16324 (2022).
Y. Wu et al., REMAP: Reliable Mechanical Energy Autonomous Platform for Monitoring Rotating Equipment, arXiv:2503.07462 (2024).
EnOcean Product Portfolio, https://www.enocean.com/en/enocean_modules
Silicon Labs, xG22E SoCs for Energy Harvesting IoT, https://www.silabs.com/blog/simplifying-ambient-iot-with-xg22e
Z. Jiang et al., REHSense: RF Energy Harvesting for Wireless Sensing, arXiv:2409.00086 (2024).