Most laser sensors time light echoes and work out speed from snapshots. Aeva's chip sends a continuous beam and reads the shift in the light, so every point carries its speed.
The company · 2 · What exactly does it sell?
Engineering fees are three fifths of sales
Each $1 of sales, first half of 2026
60%40%
Engineering feesSensors sold
Fees for adapting the sensor to each customer are paid once; sensor sales only grow when vehicles reach production.
The company · 3 · How does it make money?
Paid to develop now, per sensor later
$18.1Mrevenue in 2025
$127M to $142Myearly running costs since 2023
About $30Mcash used last quarter
Sales do not yet cover costs, so it raises money by selling new shares every few months.
Is it a good business? · 4 · What gives it an edge?
Once picked, a supplier is locked in for years
5 to 7 years
a won vehicle slot
Daimler Truck
self-driving trucks
Top-10 carmaker
a global programme won
Swapping a sensor means retesting the vehicle. Winning the slot is the edge, but only production turns it into money.
Is it a good business? · 5 · What is the catch?
Cheaper sensors and cameras already ship
Hesai, Ousterecho sensors keep getting cheaper
Mobileyestopped its own continuous-beam sensor in 2024
Aeva's answer is to sell what echo sensors cannot: speed for every point.
Is it a good business? · 6 · What does its record show?
Costs are fixed; sales have not caught up
Sales, $M
9.3
4.2
4.3
9.1
18.1
24.7
Dec 2021Dec 2022Dec 2023Dec 2024Dec 2025last 12 mo
Sales$4.3M → $24.7Msales, 2023 to the past year
Per $133.6 centskept per $1 after making it, last quarter
Sales grew several times while costs held, so the loss per dollar shrinks. It is still far from covering them.
Is it a good business? · 7 · How much room is there to grow?
Heavy trucks are the clearest market
About 300,000heavy trucks sold a year in North America
Bendixchose Aeva for its next safety system
Nobody has said how many trucks will carry the sensor, or at what price.
Is it a good business? · 8 · Who is in charge?
The founders still run it; the finance chief left
2016two former Apple engineers founded it
September 2026the finance chief of six years left
About 37%held by insiders, mostly the fund Sylebra; they only sell
Soroush Salehian is chief executive and Mina Rezk technology chief and chairman.
The story right now · 9 · Why do self-driving trucks need Aeva?
A truck that drives itself must judge speed far ahead
Daimler Truckexclusive wording for its self-driving trucks
About 300,000heavy trucks a year
Bendixa truck safety system
A heavy truck needs to see a long way ahead and know how fast everything is moving, to stop in time. Aeva's sensor reports speed directly.
The story right now · 10 · When does a design win pay?
A win pays only when vehicles reach production
A carmaker picks AevaA design win.
Aeva adapts the sensorPaid an engineering fee.
Years of testingNo sensor sales yet.
Production startsSensors sold per vehicle.
None of its design wins is in production yet, so the wins are promises, not sales.
The story right now · 11 · What changed lately?
Sales nearly doubled; the yearly forecast went unanswered
$6.7Msales, up 86%
$10.9Mwhat the last quarter must reach
About $30Mcash used in the quarter
About 51%below the 52-week high
Mixed quarter
Asked whether it is on track for $30M to $36M this year, management did not say.
The story right now · 12 · What is everyone arguing about?
Can growth leap to meet its own promise?
Many investors think
Design wins prove Aeva is taking off
We think
Growth must leap to meet its own promise
June's growth was 11.3% against the 91% to 157% the second half needs.
The price and our view · 13 · What is a share worth?
No single value here: it is priced as a bet
0.8×if it works
0%kept if it fails
Multiples of the company's value when the bet was last struck, 30 Sep 2026.
We put no single value on a share here, because the business does not yet earn a lasting profit. Instead we ask what the shares pay if the plan works, and what is left if it fails.
The price and our view · 14 · How could we be wrong?
More shares, and customers nobody names
More new sharesBiggest
It raises money every few months.
Insiders sellingShowing now
None buying.
Unnamed customersUnknown
Three made 64% of 2025 sales.
A short-seller's reportWatch
Questions the deals.
The new data-centre deals carry no volume or price yet.
The price and our view · 15 · How will we know?
The fourth quarter must reach $10.9M
Fourth-quarter sales of $10.9M or moreGood sign
A cut range, or selling at a loss againBad sign
Daimler's trucks reach productionWe turn positive
MAR 20274
Fourth-quarter 2026 results and 2027 outlook, 4 March. Companies report every three months; each report is a test of the story.
Our view moves on these business events, not on a price line, since we put no single value on a share here.
The price and our view · 16 · So, what do we think?
A real sensor edge, priced as if every win already ships
Priced for successthe kind of bet this is
UnattractiveMedium conviction
Aeva's laser sensor sees how fast every point is moving, which rivals cannot. None of its won programmes is in production, and it keeps raising money. Unattractive.
↑ ↓ to move · C for contents
What it does
More about the company
Who it is
A self-driving truck needs to know where everything around it is heading. Most laser sensors, called lidar, fire pulses of light and time the echo. That gives distance, but speed has to be worked out by comparing one snapshot with the next.
Aeva's sensor, built on a chip, sends a continuous beam and reads the tiny shift in the light that bounces back, so every point arrives with its speed attached.
How it makes money
Today it earns two kinds of money: sensors sold in small numbers, and fees from partners for engineering work to fit them, $18.1M in all in 2025. The mix swings its margin from quarter to quarter. The large, repeating money arrives only when a won programme reaches production.
Growth and profit
What growth costs
Aeva's research and running costs barely move, so each extra dollar of sales shrinks the loss: costs held flat while sales grew about fivefold since 2023. The catch is scale. Sales are still a small fraction of those costs, and the last year added more office and stock-pay spending as it builds the new data-centre business.
Its record
The numbers
Is it growing?Fast
Sales grew 63% over the twelve months to Sep 2026, slower than the 99% in the year to Dec 2025.
Is it profitable?Losing
Operating margin is −559%, up from −34× in the year to Dec 2023.
Does growth pay off?Pays off
Running costs take 590¢ of each sales dollar, down from 1162¢ in the year to Dec 2021.
Does it earn good returns?Poor
Shareholders' money lost 354% a year, down from −42% in the year to Dec 2021.
How long can it keep going?Limited
It used $114M of cash over the last twelve months, so the $148M it holds lasts about 1.3 years at that pace.
Is the balance sheet safe?Net cash
It holds $148M of cash against $97.0M of debt, $50.5M more cash than it owes.
Are shares being diluted?Shrinking
The share count fell 64% in 4.8 years, and pay in shares costs 140% of sales.
Why it wins
The evidence
A carmaker approves a sensor as part of a whole vehicle, after years of testing. Once a supplier wins that slot, rivals are locked out for five to seven years or more, because swapping means running the tests again.
Aeva holds exclusive-supplier wording on two of its biggest programmes: Daimler Truck's self-driving trucks, where early samples have passed on-road tests, and a top-10 European carmaker's global programme aimed at production in 2028. The lock only pays once production starts.
Who could beat it
The rivals in full
Both alternatives already ship. Pulse sensors keep getting cheaper: Mobileye stopped building its own continuous-beam sensor in 2024 and gave falling pulse-sensor costs as one reason. Tesla uses cameras alone, with no laser of any kind. Aeva's answer is to sell where directly measured speed matters most, such as long-distance trucking.
The sensor itself: an OEM chooses a cheaper, higher-volume ToF unit instead of Aeva's FMCW unit where instantaneous per-point velocity is not required to close the safety case.
LiDAR of any architecture: the whole sensing modality is left out of the vehicle rather than substituted sensor-for-sensor.
How big it could get
The market and who has signed
The market
The clearest market is heavy trucks: 208,155 of the largest class were sold in the US in 2025, a down year, and about 300,000 a year across North America. Bendix, whose safety system is on most of them, has chosen Aeva for its next version. Nobody has said how many trucks will carry it, or when, so the market cannot honestly be sized.
Who has signed
None of the three deals that matter comes with a volume or a length. Daimler Truck and the European carmaker are design wins: real choices, paid only if the vehicles reach production. A deal to put Aeva's laser light source inside a large cloud company's data-centre equipment is a development agreement, not an order.
Who runs it
Leaders, pay and the cash
Who runs it
Two former Apple engineers, Soroush Salehian and Mina Rezk, founded Aeva in 2016 and still run it, as chief executive and as technology chief and chairman. The finance chief of six years left in September; the company's controller stands in until a permanent hire. Insiders hold about 37%, mostly the hedge fund Sylebra, whose founder sits on the board and also provides the standby funding. The founders' bonus is at the board's discretion, and most of their stock simply vests with time.
What they did with the money
It raised money by selling shares, as it has every few months since 2023. A $115M sale in June lifted the share count 11.5% in one quarter. The $177.9M it held at June funds about 17 months at that quarter's spending, and its finance chief of six years left in September.
The last report
The latest report
Aeva's sales nearly doubled to a record $6.7M, but the yearly guide went unanswered
What it did to the thesis Mixed quarter
The growth and margin test passed, which is why the case against the shares weakened a little. But the yearly guide is unconfirmed, cash keeps falling and the new deals are talks with no volume or price. The view holds at Unattractive.
The report
Sales grew 86% to a record $6.7M and Aeva kept 33.6 cents of each sales dollar after the cost of making it, a second positive quarter. It booked its first sales of parts for data-center optical links.
Asked on the call whether it is still on track for $30M to $36M this year, management did not say. Nine months came to $19.1M, so the last quarter must reach at least $10.9M, and cash fell about $30M in the quarter.
What's happening now
The background
The shares are down about 51% from their 52-week high, even after a 28.8% jump on the June results and the data-centre announcement. The fall has not made them cheap.
The test is close. On 4 November Aeva reports the September quarter, the first chance to see whether growth is returning to the pace its own promise needs. The $125M facility it can draw on by selling stock expires the same month.
Where we disagree
Both sides in full
The market
The design wins and the new data-centre deal prove Aeva is taking off.
We believe
The nearer test is arithmetic: June's 11.3% growth against the 91% to 157% the second half needs.
The argument
Six analysts rate the shares a buy and two a hold, and all ten questions on the June call were about the new opportunities. None asked about the growth gap, the cash use or the new shares.
What would prove us wrong is plain: September-quarter sales growing back toward the promised pace.
What it's worth
How we value it
No fair value: the analysts' forecasts show no lasting operating profit after stock pay, so a value built on earnings would rest on a guess; the bet prices it instead.
The bet
Even if this works, the shares look expensive. The best case we can defend is two analysts' sales forecast for 2028, several times this year's sales, and valued like a working lidar business that is worth less than today's price. Getting there needs a data-centre deal the size management describes, or several car and truck programmes reaching production together, and none has a signed volume yet.
If it fails, shareholders are left mostly with cash that shrinks every quarter, with convertible debt ahead of them.
The bet
Seen as a bet
Even if this works, the shares look expensive. The best case we can defend is two analysts' sales forecast for 2028, several times this year's sales, and valued like a working lidar business that is worth less than today's price. Getting there needs a data-centre deal the size management describes, or several car and truck programmes reaching production together, and none has a signed volume yet.
If it fails, shareholders are left mostly with cash that shrinks every quarter, with convertible debt ahead of them.
What could go wrong
Every risk
The risks
More new shares
Aeva has raised outside money every few months since 2023, and about 17 months of cash makes another raise likely.
Insiders selling, none buying
58 open-market sales and no purchases in a year, including about $9.7M by the two founders on two days in August.
Few, unnamed customers
Two customers made up 40% and 19% of first-half sales, and neither is named.
A short-seller's report
Fugazi Research called Aeva uninvestable in August, and every figure it used matches Aeva's own filings.
What we still do not know
Who the customers are. Three made up 64% of 2025 sales, and none is named.
How many of the roughly 300,000 heavy trucks sold in North America each year will carry the new Bendix system, and from when.
Whether the data-centre deal becomes an order. Management spoke of hundreds of millions of dollars a year on the call; no filing says so.
What changes our view
Every route to a new view
Every route to a new view
→ Attractive
Daimler's self-driving trucks reach production, or the data-centre deal becomes a supply contract with a disclosed price and volume. The shares fall to about 16.5 times sales, where the other listed laser-sensor makers trade.
Research reopens
Full-year 2026 sales land inside the promised $30M to $36M, with Aeva still making a profit on what it sells.
The next test
It supports the case if
Q4 revenue at least $10.9 million with gross margin positive and the guide met; or the optical-connectivity relationship converts to a priced supply contract.
It pressures the case if
Q4 revenue short of $10.9 million, a narrowed or withdrawn range, gross margin turning negative, or a large share sale funding the cash burn.
The verdict
Scores and how we weighed it
The scores
Growth
3 of 4
Business quality
3 of 4
Financial position
2 of 4
Execution record
3 of 4
Per share capture
2 of 4
Potential
2 of 4
Risk profile
3 of 4
Valuation
1 of 4
Evidence quality
3 of 4
How we weighed it
ValueMixed
No fair value is drawn: the forecasts show no lasting profit after stock pay, so the bet prices it.
BusinessMixed
A mixed business: a unique sensor, with no programme in production.
GrowthPositive
Growth and margin passed the test we set last quarter.
RisksNegative
More new shares are coming, insiders keep cashing out and its customers are few and unnamed.
RivalsNegative
Near the top of its rivals on sales and the dearest on what they own.
The groupMixed
Lidar names move together on robotics and car news.
Latest newsMixed
The last report passed our test, but cash keeps falling and the new deals are only talks.
In plain words
Design win
A carmaker choosing a supplier's part for a future vehicle. It promises sales only once that vehicle is built, often years later, and the number built is not fixed.