Arbe Robotics published its second-quarter report on 6 August 2026. We are writing about it today.
And this is a company that has to be measured differently from everything else we covered this week.
What was reported
| Quarter | Year ago | |
|---|---|---|
| Revenue | 0.7 | 0.3 |
| Backlog at 30 Jun 2026 | 1.0 | |
| Cash and deposits | ~42 |
In millions of dollars
Why revenue is not the measure
At this stage, cash is the story
$0.7 million of quarterly revenue is not a business. It is proof of feasibility.
Arbe develops high-resolution imaging radar for autonomous vehicles - technology sold to carmakers and tier-one suppliers, and in that industry the distance between a signature and mass production is measured in years.
So the number that matters is how long the company can survive until that happens:
- Cash and deposits: roughly $42 million
- Target burn rate: below $7 million per quarter
42 divided by 7 is six quarters - roughly a year and a half, if the target is met and no further financing arrives.
And management phrases it itself: the current balance sheet, together with revenue growth and a targeted reduction in burn, extends the runway.
What is moving
The company reports an order from an L4 autonomy OEM customer - meaning full autonomous driving under defined conditions, not merely driver assistance.
And a collaboration with HiRain, a Chinese tier-one supplier, which per the report is competing in meaningful RFIs and RFQs.
The company also cites sales to automotive tier-ones, and radar system sales for defense and civilian programs.
And that detail is interesting: Arbe too, like Innoviz which reported a day earlier, mentions the defense market as an additional channel alongside automotive. Two Israeli sensor companies looking in the same direction.
My Angle
A personal opinion of Ilan Abramov - not advice, not a recommendation
I want to be direct about how this report should be read.
A company with $0.7 million in quarterly revenue and a $1 million backlog is not measured on profit, growth or margin. It is measured on one question: how long does it have before it needs to raise again.
And the answer here is roughly a year and a half - $42 million of cash against a burn target of under $7 million a quarter. That is a reasonable runway, but it is conditional on the target being met.
And what I look for at a company like this is not revenue but milestones: whether the L4 customer order becomes a production program, and whether HiRain wins the tenders it is competing in. Those are the events that change the numbers, not the quarter itself.
And what interests me in the wider context is that both Arbe and Innoviz are talking about the defense market this week. Two Israeli sensor companies built for the automotive market, discovering that defense orders arrive faster. That is not coincidence - it is what happens when automotive timetables slip and defense budgets grow.
