Tutorials Video

Eric Fossum: The Sensor in Every Phone

By Infinite Creation · Profile

1957-

No freely licensed image known

At a Glance

Eric R. Fossum (born 1957) invented the CMOS active-pixel sensor with intra-pixel charge transfer at NASA's Jet Propulsion Laboratory in the early 1990s, seeking to miniaturise cameras for spacecraft. It can be manufactured in standard CMOS processes alongside its own processing circuitry, and it is the sensor in essentially every smartphone.

Born 1957, United States
Anchor year: 1993, the CMOS active-pixel sensor
Strand: overlooked, billions of units, no public profile

Career

The problem as it stood

CCDs produce excellent images and need specialised fabrication, multiple supply voltages, substantial power, and separate chips for processing. For a spacecraft, or later a phone. That is size, weight, power and cost in all the wrong places.

What he actually did

Fossum's team put an amplifier in every pixel and used intra-pixel charge transfer to a sense node, giving each pixel its own active readout while retaining the low noise that earlier passive CMOS sensors lacked. The result is manufacturable in a standard CMOS process.

How it worked

A CCD marches charge across the whole array to one amplifier. An active-pixel sensor amplifies at each pixel and reads them by address, like memory. That means no transfer losses, low power, random access, very fast readout, and. Decisively, the ability to put the sensor and its processing on the same die in an ordinary fab.

What it made possible

Cameras cheap and small enough to put in everything. Smartphones, webcams, cars, doorbells, endoscopes. Almost all video now produced in the world, including most of what is watched through the delivery chain in the previous wave, originates on one of these.

What happened to him

He commercialised the technology through Photobit, later acquired by Micron, and became a professor at Dartmouth. He received the Queen Elizabeth Prize for Engineering in 2017 with Boyle, Smith and Tompsett: an award that credited all four.

Where the credit landed

Fossum is well credited within the field and unknown outside it, despite his invention shipping in billions of units annually. The foundational patent itself names three inventors, Fossum alongside Sunetra Mendis and Sabrina Kemeny, and this page credited only the first of them until August 2026. The Queen Elizabeth Prize is notable for doing what the Nobel did not: crediting the imaging lineage collectively, across Boyle, Smith, Tompsett and Fossum, rather than selecting two names from it.

Known For

An amplifier in every pixel means no transfer losses and random access.
Standard CMOS fabrication means sensor and processing share a die.
The Queen Elizabeth Prize credited all four CCD/CMOS figures, the Nobel credited two.

Knowledge Base

What You'll Learn

The last link in capture: a camera cheap enough to put in everything.

Rolling shutter, and where it comes from

Because a CMOS sensor is read out row by row rather than all at once, different rows are exposed at slightly different times. Fast horizontal motion therefore appears skewed, and a propeller can look bent, the rolling shutter effect. Global shutter designs avoid it at the cost of complexity and pixel area. It is a direct visible consequence of the readout architecture that made these sensors cheap.

Where This Fits

The CMOS sensor is the capture link in almost every modern camera, the starting point of A History of Broadcast Technology.

FAQ

Q: What is a CMOS active-pixel sensor?
A: An image sensor with an amplifier in each pixel, read by address rather than by shifting charge across the array. It can be made in standard CMOS processes alongside its own processing circuitry, making it small, cheap and low-power.

Q: Why do phone videos sometimes look skewed?
A: Rolling shutter. CMOS sensors are read row by row, so rows are exposed at slightly different times and fast motion appears distorted.

Vocabulary

CMOS sensor. An image sensor built with the same manufacturing process as ordinary computer chips. Allows the sensor and all its supporting circuits to sit together on one piece of silicon.

Active pixel. A sensor cell containing its own amplifier, so the signal is strengthened where it is generated rather than after being moved across the chip. Reduces the noise that accumulates during transfer.

Intra-pixel charge transfer. Moving accumulated charge within a single cell before it is read out. Improves noise performance substantially, and is what made the design competitive with the sensor types that preceded it.

Camera on a chip. Integrating the sensor, timing, conversion and control circuitry onto one component. Reduced cameras from assemblies of separate parts to something small enough to fit inside a telephone.

Sources

The patent was checked against the record, and it corrected this page. Checked 20 August 2026.

US Patent 5,471,515, Active Pixel Sensor with Intra-Pixel Charge Transfer. Filed 28 January 1994 and granted 28 November 1995, assigned to the California Institute of Technology. The record names three inventors, Eric R. Fossum, Sunetra Mendis and Sabrina E. Kemeny. This page previously credited Fossum alone, and was corrected once the patent was read.