The cell looks perfect.
On a battery line somewhere, a lithium-ion cell rolls past a station the size of a filing cabinet. No clamp opens it. No blade cuts it. A pulse of ultrasound passes through it and, in the time it takes you to read this sentence, a verdict appears on a screen: good, or hiding something. That cabinet is made by Titan Advanced Energy Solutions, and it has just done what the rest of the line cannot - looked inside a sealed battery without breaking it.
Titan AES is a 43-person company in Salem, Massachusetts - a town historically better at finding witches than gigafactory defects. Its product, IonSight, gives every cell on a production line the equivalent of a medical ultrasound. The pitch is unglamorous and enormous at once: most of the world's batteries ship after only a handful of their siblings were ever inspected. Titan thinks every single one deserves a look.
Batteries keep their secrets well
A finished lithium-ion cell is a sealed metal can. Inside, dozens of microscopically thin layers are wound or stacked with margins for error measured in fractions of a millimeter. A wrinkle in a separator, a dry patch in the electrode, a stray metal particle - any of these can pass every electrical test and still become a field failure, a swollen phone, or a headline-grade EV recall years later.
The legacy options for looking inside were never built for volume. CT scanning is slow and expensive, so factories sample a few cells per batch and hope the rest behave. Destructive teardown answers the question by killing the patient. Electrical testing measures how a cell performs, not how it was built. The result is an industry that ships millions of cells while having truly inspected very few of them.
Two founders, one strange idea
In 2016, Shawn Murphy and Sean O'Day made a bet that sounded faintly absurd: the same physics behind a sonogram - sound waves bouncing off internal structures - could map the guts of a battery cell at the speed a factory actually runs. Ultrasound is cheap, fast, and harmless to the thing it inspects. It had simply never been pointed at batteries with enough resolution or enough software to make sense of the echoes.
Murphy runs the company as CEO and CTO; O'Day serves as president and chief commercial officer. The wager was that the hard part was not the sound - it was teaching a machine to read it. Every defect leaves a different acoustic fingerprint, and turning a blur of returning waves into “layer misalignment, row 12” is an AI problem dressed as a hardware one.
Shawn D. Murphy
CEO & CTO. The technical half of the bet - convinced that sound could out-see a CT scanner.
Sean O'Day
President & CCO. The half who had to convince gigafactories that “inspect everything” was a budget line, not a fantasy.
The two-person physics argument that turned into a company with the U.S. Department of Energy on speed dial.
IonSight: a stethoscope for the assembly line
IonSight Digital Teardown is the flagship. It pairs high-resolution ultrasound sensing with industrial AI to image a cell's internal morphology and flag what shouldn't be there: layer misalignment, moisture contamination, dry zones, separator wrinkling, electrode buckling, foreign particles. It collects more than 500,000 data points from a single cell, then hands a verdict back to the line before the next cell arrives.
The catch with inspection has always been speed. A test that slows the line is a test nobody buys. So Titan built Atlas, the in-line model that keeps pace at 55-plus cells per minute on a 12 GWh factory line. There is also Scorpion, a separate diagnostic that reads a lithium-ion battery's condition in seconds using hardware, cloud software, and machine learning - the same listening trick, aimed at batteries already out in the world.
What a manufacturer gets out of all this is less a gadget than a feedback loop. Defects get classified the moment they appear, so a process engineer can trace a contamination problem back to the coating step that caused it instead of guessing across a week of production. Suppliers can be qualified on inbound cells before they ever enter the line. Yield problems that used to take weeks to diagnose become questions with same-shift answers. The cabinet is the visible part; the real product is the intelligence that comes out of it.
The Titan timeline
Who's listening
A clever idea is easy. Getting industrial giants to bolt your cabinet onto their line is the hard part - and that is where Titan's claim stops being a slide deck. Honeywell folded IonSight into its Battery Manufacturing Excellence Program. Siemens signed on to chase scrap reduction and quality. Schneider Electric is both a partner and, through SE Ventures, an investor. Cell maker Navitas Systems put the technology into a real deployment. The U.S. Department of Energy picked Titan to help reshore advanced battery manufacturing.
The argument for ultrasound over the incumbents is a numbers argument, so here are the numbers.
Cells actually inspected, per approach
Legacy methods sample; Titan's pitch is total coverage. Bars are illustrative of the coverage gap, not a vendor benchmark. Throughput claim: Atlas handles 55+ cells/min on a 12 GWh line; 500,000+ data points captured per cell.
Better batteries, fewer surprises
Titan's stated mission - “Battery Intelligence Based on Sound Science” - is a pun that happens to be a strategy. The company wants every cell inspected before it powers a phone, a car, or a slice of the grid. Catch the defect on the line and you prevent the swollen battery, the warranty claim, the recall, the fire. You also speed up factory ramps, because a manufacturer who can see inside a cell can find the root cause of a yield problem in hours instead of guessing for weeks.
There is a climate angle hiding in the quality angle. Every defective cell caught early is material not wasted, a battery pack not scrapped, an EV or storage project not delayed. Quality control, it turns out, is quietly one of the more boring and effective forms of decarbonization.
It is worth noting that Titan is not alone in pointing sound at batteries - Liminal Insights and a handful of others share the thesis, and CT and electrical testing are not going away. But the company has spent the better part of a decade on a single, narrow question: how do you see inside every cell, fast enough and cheaply enough that “every” is realistic? That focus is the moat. The competitors validate the market; the partner logos suggest who is winning the early innings.
The cell still looks perfect.
The world is about to make staggeringly more batteries than it ever has - for cars, for grids, for the slow electrification of nearly everything. Each one is a sealed can with the potential to be flawless or quietly defective, and the difference is invisible from the outside. The old answer was to inspect a few and trust the rest. That answer scales badly when the stakes are measured in recalls and fires.
So picture the line again. The cell rolls past the filing-cabinet-sized station. The ultrasound pulses through. This time we know what the screen is doing - it is reading a battery's secrets out loud, before the battery gets the chance to keep them. The cell still looks perfect from the outside. Titan AES is the reason that no longer has to be a guess.