The Hidden Records of Tesla Autopilot Fatalities
A fatal collision in Ohio reveals how Tesla uses regulatory redactions to obscure the performance of its driver-assist systems in real-world crashes.
By LNGFRM Team
Published September 3, 2026
“Fool’s errand.” That’s the dismissive decree from Elon Musk regarding lidar technology, a pronouncement that sent ripples through the autonomous vehicle industry and cemented Tesla’s camera-only approach.
Yet, just as the tide of innovation often flows against prevailing currents, a former Tesla engineer, Eric Aguilar, has not only rejected his old boss’s dogma but has staked his entire professional future on the very sensor Musk so readily cast aside.
Aguilar, the cofounder and CEO of Omnitron Sensors, operates on a profound conviction: the future of self-driving cars and, crucially, humanoid robots, hinges on lidar’s ability to see the world with unparalleled precision.
Aguilar’s journey is rich with a certain poetic irony.
He was instrumental in the rollout of Tesla’s Model 3, a vehicle emblematic of Musk’s vision.
But by 2018, the paths diverged.
Armed with two decades of experience in sensor technology, including stints at Google X’s ambitious drone delivery project and the now-defunct robotaxi venture Argo AI, Aguilar understood the allure and the limitations of traditional lidar.
He recognized Musk’s misgivings were not unfounded; early lidar systems were indeed cumbersome, costly, and notoriously fragile.
They were, in essence, a mechanical marvel ill-suited for the rigors of commercial deployment.
The problem, as Aguilar meticulously explains, lies in lidar’s very architecture.
Imagine a miniature disco ball, its laser beam sweeping an environment, measuring distances through reflected light.
To achieve greater range and fidelity, you typically need larger lasers and, consequently, larger, heavier mirrors.
This design philosophy leads to a sensor riddled with numerous moving parts, susceptible to wear and tear, and demanding frequent, costly recalibration.
Automakers, understandably, recoil at the prospect of replacing critical sensors every two to three months.
“When you talk to OEMs — Mercedes, Volvo, BMW, GM — they don’t want lidar to have to be replaced every two or three months,” Aguilar candidly shared, echoing the frustrations he witnessed firsthand.
This is where Omnitron Sensors enters the fray, not merely to improve lidar, but to fundamentally reimagine its construction.
Aguilar, an electrical engineer by training, envisions a future where lidar is not a delicate, expensive instrument but a robust, mass-produced component.
His solution: harness the power of silicon and the semiconductor manufacturing process.
By crafting lidar components on silicon wafers, Omnitron aims to drastically reduce the number of moving parts, enhance durability, and achieve nanometer-level precision.
This shift from bespoke mechanical assembly to scalable chip fabrication promises to slash costs from the staggering $120,000 per unit of old to mere “hundreds of dollars,” a price point Aguilar believes will finally “unlock the market.” But why is this technological leap so critical?
For autonomous vehicles, lidar isn’t just an auxiliary sensor; it’s a vital safeguard against the unpredictable chaos of the real world.
Cameras, while powerful, grapple with low-light conditions and can be catastrophically confused by shadows.
Is that a real object, or just an optical illusion cast by the sun?
Lidar, impervious to such ambiguities, provides an indisputable, three-dimensional map of the environment, crucial for navigating those “rare edge cases” that haunt the developers of self-driving systems.
Beyond the asphalt, Aguilar sees an even more compelling need for lidar in the burgeoning field of humanoid robots.
As these mechanical beings step out of laboratories and into homes, factories, and public spaces, their ability to perceive and interact safely with their surroundings becomes paramount.
Just as humans rely on a complex interplay of senses to gauge force and distance, robots will need lidar to understand the depth of an object, the contours of a room, or the precise distance to a human child.
The bar for integrating robots into the human experience is “much higher,” Aguilar stresses, particularly when considering scenarios like a robot holding a baby.
Robust perception isn’t just an enhancement; it’s a prerequisite for trust and safety.
Aguilar’s thesis resonates with a growing chorus of voices in the autonomy sector, all underscoring the non-negotiable role of safety.
While Musk’s vision prioritizes a lean, camera-centric approach, Aguilar’s work at Omnitron suggests that true, uncompromised autonomy across diverse environments—from bustling city streets to intricate indoor settings—demands the comprehensive spatial understanding that only lidar can reliably provide.
His bet against the titan of Tesla isn’t merely about technology; it’s about the very foundation upon which a safe and widely accepted autonomous future will be built.
If Aguilar succeeds in making lidar affordable, durable, and ubiquitous, he won’t just be proving Musk wrong; he’ll be paving the way for a generation of robots that can truly see, and therefore, truly operate, in our complex world.
A fatal collision in Ohio reveals how Tesla uses regulatory redactions to obscure the performance of its driver-assist systems in real-world crashes.
New York City lawmakers are moving to permanently prohibit armed police robots, transforming an internal NYPD policy into a binding law and preempting a future where lethal force is delivered by machine.
Former Tesla engineer Eric Aguilar is challenging Elon Musk’s camera-only approach to autonomy with his startup, Omnitron Sensors. His company is developing durable, silicon-based lidar to make self-driving vehicles and robots safer and more affordable. Aguilar believes this technology is critical for robust perception in the robotic revolution.
