Picture a sensor on every case in a distribution center. Now picture the battery inside each one.
‍

Energy harvesting is the process of capturing small amounts of energy already present in the environment, such as radio waves, light, heat, or motion, and converting it into electricity for low-power electronics. It’s what makes battery-free technology possible: sensors that keep sensing and reporting with no battery to replace.
‍

That matters more than it sounds. With batteries, somebody has to buy them, install them, track which ones are dying, find the devices that went quiet, swap them out, and dispose of the old ones. At a few hundred devices, that's a maintenance task. At a few million, it's a second operation running on top of the first.
‍

Here's how battery-free sensing works, what it can realistically power, and why it's becoming the foundation for the next wave of supply chain AI.
‍

How Does Battery-Free Sensing Work? Capture, Convert, Spend

Every energy harvesting system does three things.
‍

Capture. The device collects energy from its surroundings. The IEEE describes the basic principle as collecting ambient energy and converting it for use where wired power isn't available or battery replacement isn't practical.
‍

Convert. A component called a transducer turns that input into electrical current. A photovoltaic cell does it with light. An antenna and rectifier do it with radio waves. Thermoelectric and piezoelectric materials do it with temperature differences and vibration.
‍

Spend carefully. The power collected is tiny, so the device budgets it. It sleeps most of the time, wakes briefly to sense and transmit, then goes back to sleep. Power management circuitry holds small amounts of harvested energy and releases it when there's enough to do useful work.
‍

That last step is the one people underestimate. Energy harvesting doesn't make a device powerful. It makes a device self-sufficient, as long as the job is small and repeatable.
‍

Where the Energy Comes From

‍

Not every source works everywhere. The right one depends on where the sensor actually lives.

Light is the most familiar. Small solar cells work well on exposed surfaces, but a sensor inside a sealed carton or at the bottom of a pallet stack won't see much of it.
‍

Thermal energy needs a temperature difference, not just warmth. A gradient between two surfaces can generate power. A product sitting at one uniform temperature gives the device very little to work with.
‍

Motion and vibration suit things that move, like equipment, vehicles, and assets in transit. They go quiet the moment the asset sits still.
‍

Radio frequency (RF) energy is the one that matters most for supply chains. Radio waves from Bluetooth, Wi-Fi, cellular, and other wireless networks already fill the places goods move: warehouses, trailers, stores, back rooms. You can't see them, but they're consistent, and they don't depend on lighting, temperature, or whether the item is moving.
‍

What Can Battery-Free Sensors Actually Power?

Harvested energy is measured in microwatts to milliwatts. That won't run a handheld scanner, a screen, or a refrigeration unit. It isn't supposed to.
‍

What it can power is exactly what supply chain sensing needs: a brief measurement and a short wireless transmission, repeated over and over. Is the item here? Where is it? How long has it been sitting? What temperature is it? Did it move when it shouldn't have?
‍

None of those questions takes much power on its own. Answered continuously, across every item in a network, they add up to a picture no periodic scan can match.
‍

Energy harvesting doesn't make a sensor powerful. It makes it small enough, cheap enough, and maintenance-free enough to live on the item itself.

Why Removing the Battery Changes the Math

It's tempting to treat battery-free as a convenience feature. It's closer to a change in what's economically possible.
‍

Scale stops being a maintenance problem. Every battery-powered sensor comes with its own service clock. Battery-less sensors don't, so a deployment can grow from thousands of items to millions without a matching jump in service labor.
‍

The form factor shrinks. No battery means no battery compartment. Battery-free sensors can be manufactured in a thin, label-like form and applied directly to products or packaging instead of bolted on as separate devices.
‍

One-way use becomes viable.When there's no battery to recover, it makes sense to tag things that never come back, like a case of produce or a single outbound shipment.
‍

This is also where battery-free sensing differs from the traditional passive kind. Conventional passive tags only respond when a reader energizes them at a fixed point, like a gate or a handheld. Between those reads, they're silent. Sensors that harvest ambient RF energy can broadcast on their own, continuously, wherever there's signal to harvest.
‍

How IoT Pixels Use Energy Harvesting

Wiliot's IoT Pixels are built on this model. They're stamp-sized, battery-free Bluetooth sensors that harvest RF energy from existing sources, including Bluetooth, Wi-Fi, sub-1GHz, and cellular networks. That energy powers continuous sensing of location, temperature, humidity, motion, and light, all from a single tag.
‍

Because they don't wait for a scan, IoT Pixels typically report multiple events per minute. The most complex scan-based tracking systems may capture fewer than 10 data points on an asset across its entire journey. IoT Pixels produce that many every few minutes. And since sensing never stops, they can measure what checkpoint reads structurally can't, like how long an item has been sitting in one spot.
‍

That stream flows into the Wiliot Physical AI Platform, where AI and statistical models refine raw signals into events, predictions, and recommended actions. A temperature drift becomes an early spoilage warning through Temperature Monitoring. An idle roll cage becomes a loss alert through Reusable Asset Tracking. A count that doesn't match becomes a correction in Inventory Intelligence.
‍

Energy Harvesting Is the Foundation of Physical AI

Supply chain AI is moving quickly toward agents that act on their own: reordering, rerouting, escalating exceptions. But AI can only reason over what it can observe, and most of the physical world still goes unobserved between scans.
‍

The solution is putting a sensor on everything that matters. That's only realistic if the sensors cost very little, last as long as the item, and never need a battery change.
‍

Energy harvesting is what makes that math work. It's the reason a continuous, item-level sensing layer is possible at all, and that sensing layer is what lets AI reason over reality instead of over records.
‍

The battery was always the bottleneck. Take it away, and the physical world becomes something AI can finally see.
‍

Curious what continuous, battery-free sensing could show you about your own operation? Talk to our team about putting Wiliot to work in your supply chain.

‍