News
Mobileye CTO on Building Autonomous Driving Systems
Professor Amnon Shashua, co-founder and CTO of Mobileye, gave a talk recently explaining how his company builds its autonomous driving systems. One point he emphasizes is that identifying other vehicles and people in the vicinity of the car is the easy part. What’s difficult to achieve is controlling the car so that it stays within the limits of the roadway, especially when road markings, curbs, and other indications of where the physical edges of the permissible driving space are lacking.
Shashua breaks the process into three main components — sensing, road experience management, and driving policy, which he calls “planning.” The first collates all the data coming into the system from cameras, radar, and, in some cases, Lidar. The second phase creates an environmental model that orients the car and determines its place on the road. The third phase uses all the prior information to determine all the available pathways forward and which is preferred.
Next, he goes on to speak in detail about what he calls “deep learning.” Shashua says despite its current popularity, there is a great deal of hype surrounding it. He explains why some aspects of deep learning fail to give the zero failure results required to make accurate autonomous driving possible. In particular, he shows why what he calls the “end to end approach” does not live up to its promise.
He prefers what he calls “semantic abstraction,” a process that breaks down the tasks the system must perform into discrete subsystems, each with its own function to perform. He feels the latter is more accurate and gives the fewest false positives or false negatives. The end to end system does not require image annotations, but Shashua says Mobileye has 600 people on staff annotating images and expects to have 1,000 or more shortly.
What is important, Shushua maintains, is dealing with what he calls “corner cases.” These are the anomalies that happen in regular driving that are rare, unexpected, or unusual — like a construction vehicle moving slowly near a highway. It is those corner cases that require the most sophisticated technological solutions and the types of situations that end to end systems do not handle well.
Watch the entire 30 minute video for yourself to learn more about some of the elements that go into Tesla’s Autopilot technology.
Source: Mobileye via YouTube
Tesla became the first company to pass the United States government’s new Advanced Driver Assistance Systems (ADAS) testing with the Model Y, completing each of the new tests with a passing performance.
In a landmark announcement on May 7, the National Highway Traffic Safety Administration (NHTSA) declared the 2026 Tesla Model Y the first vehicle to pass its newly ADAS benchmark under the New Car Assessment Program (NCAP).
Model Y vehicles manufactured on or after November 12, 2025, met rigorous pass/fail criteria for four newly added tests—pedestrian automatic emergency braking, lane keeping assistance, blind spot warning, and blind spot intervention—while also satisfying the program’s original four ADAS requirements: forward collision warning, crash imminent braking, dynamic brake support, and lane departure warning.
The NHTSA has just officially announced that the 2026 @Tesla Model Y is the first vehicle model to pass the agency’s new advanced driver assistance system tests.
2026 Tesla Model Y vehicles, manufactured on or after Nov. 12, 2025, successfully met the new criteria for four… pic.twitter.com/as8x1OsSL5
— Sawyer Merritt (@SawyerMerritt) May 7, 2026
NHTSA administration Jonathan Morrison hailed the achievement as a milestone:
“Today’s announcement marks a significant step forward in our efforts to provide consumers with the most comprehensive safety ratings ever. By successfully passing these new tests, the 2026 Tesla Model Y demonstrates the lifesaving potential of driver assistance technologies and sets a high bar for the industry. We hope to see many more manufacturers develop vehicles that can meet these requirements.”
The updates to NCAP, finalized in late 2024 and effective for 2026 models, reflect growing recognition that ADAS features are no longer optional luxuries but essential tools for preventing crashes.
Pedestrian automatic emergency braking, for instance, targets one of the fastest-rising causes of roadway fatalities, while blind spot intervention and lane keeping assistance address common sources of side-swipes and run-off-road incidents. By incorporating objective, performance-based evaluations rather than mere presence of the technology, NHTSA aims to give buyers clearer data on real-world effectiveness.
This milestone arrives at a pivotal moment when vehicle autonomy is transitioning from science fiction to everyday reality.
Tesla’s Full Self-Driving (FSD) software and the impending rollout of robotaxis underscore a broader industry shift toward higher levels of automation. Yet regulators and consumers remain cautious: safety data must keep pace with technological ambition.
The Model Y’s perfect score on these ADAS benchmarks validates that current driver-assist systems—when engineered rigorously—can dramatically reduce human error, which still accounts for the vast majority of crashes.
For Tesla, the result reinforces its long-standing claim of building the safest vehicles on the road. More importantly, it signals to the entire auto sector that meeting elevated federal standards is achievable and expected.
As autonomy edges closer to Level 3 and beyond, where drivers may disengage more fully, such independent verification becomes critical. It builds public trust, informs purchasing decisions, and accelerates the development of systems that could one day eliminate tens of thousands of annual traffic deaths.
In an era when software-defined vehicles promise transformative mobility, the 2026 Model Y’s NHTSA triumph is more than a manufacturer accolade—it is a regulatory green light that autonomy’s future must be built on proven, testable safety foundations. The bar has been raised. The industry, and the roads we share, will be safer for it.
Tesla is going to fix the nearly 219,000 vehicles that it recalled due to an issue with the rearview camera with a simple software update, giving owners no need to travel to a service center to resolve the problem.
Tesla is formally recalling 218,868 U.S. vehicles after regulators discovered a software glitch that can delay the rearview camera image by up to 11 seconds when drivers shift into reverse.
The affected models include certain 2024-2025 Model 3 and Model Y, as well as 2023-2025 Model S and Model X vehicles running software version 2026.8.6 and equipped with Hardware 3 computers. The National Highway Traffic Safety Administration (NHTSA) determined the lag violates Federal Motor Vehicle Safety Standard 111 on rear visibility and could increase crash risk.
Yet this is no ordinary recall. Owners do not need to schedule a service-center visit, hand over keys, or wait for parts.
Tesla fans call for recall terminology update, but the NHTSA isn’t convinced it’s needed
Tesla identified the issue on April 10, halted further deployment of the faulty firmware the same day, and began pushing a corrective over-the-air (OTA) software update on April 11.
By the time the NHTSA posted the recall notice on May 6, more than 99.92 percent of the affected fleet had already received the fix. Tesla reports no crashes, injuries, or fatalities linked to the glitch.
The episode underscores a deeper problem with regulatory language. For decades, “recall” meant hauling a vehicle to a dealership for hardware repairs or replacements. That definition no longer fits software-defined cars. When a fix arrives wirelessly in minutes — identical to an iPhone update — the term evokes unnecessary alarm and misleads the public about the actual risk and remedy.
Elon Musk has repeatedly called for exactly this change. After earlier NHTSA actions, he stated plainly: “The terminology is outdated & inaccurate. This is a tiny over-the-air software update.” On another occasion, he added that labeling OTA fixes as recalls is “anachronistic and just flat wrong.”
The terminology is outdated & inaccurate. This is a tiny over-the-air software update. To the best of our knowledge, there have been no injuries.
— Elon Musk (@elonmusk) September 22, 2022
Musk’s point is simple: regulators must evolve their vocabulary to match the technology. Traditional recalls involve physical intervention and downtime; OTA updates do not. Retaining the old label distorts consumer perception, inflates perceived defect rates, and slows the industry’s shift to faster, safer software iteration.
Tesla’s rapid, remote remedy demonstrates the safety advantage of over-the-air capability. Problems that once required weeks of dealer appointments are now resolved in hours, often before most owners notice. As more automakers adopt software-first designs, the entire regulatory framework needs to catch up.
Updating “recall” terminology would align language with reality, reduce public confusion, and recognize that modern vehicles are no longer static hardware — they are continuously improving computers on wheels.
For the 219,000 Tesla owners involved, the process is already complete. The camera works, the car is safe, and no one left their driveway. That is the new standard — and the vocabulary should reflect it.
News
Tesla is seeing record sales rebounds in key markets globally
Tesla reported robust sales momentum in April 2026, extending a multi-month recovery in its two largest markets amid intensifying global EV competition.
Tesla is seeing record sales rebounds in key markets across the world, and as skeptics and bears of the company that builds electric powertrains rejoice on the weak registration figures that have been reported in the past, the Musk-fronted company is keen on making a comeback.
Tesla reported robust sales momentum in April 2026, extending a multi-month recovery in its two largest markets amid intensifying global EV competition.
While the company does not release official monthly global delivery figures—reserving those for quarterly reports—data from local registration and wholesale sources show significant year-over-year gains in China and several European countries, building on a turnaround from 2025’s declines.
In China, Tesla’s Shanghai Gigafactory shipped 79,478 Model 3 and Model Y vehicles in April, a 36% increase from the same month last year. The figure marks the sixth consecutive month of year-on-year growth for China-made EVs, which include both domestic sales and exports to Europe and other regions.
Although down slightly from March’s 85,670 units, the April performance underscores Tesla’s resilience against domestic rivals like BYD. Wholesale volumes from the plant have helped Tesla regain ground after softer retail figures earlier in the year, with analysts noting improved demand fueled by competitive pricing and new configurations
Europe also delivered encouraging results. Registrations—a close proxy for sales—surged in multiple countries. France posted a 112 percent jump, Sweden 111%, Denmark 102%, and Ireland 100%. The Netherlands rose 23%, while Belgium and Romania recorded gains of 47% and 53%, respectively.
These double- and triple-digit increases reflect a broader EV market recovery across the continent, where battery-electric vehicle market share climbed to 20.5% in Q1 2026 from 13.2% a year earlier. Chinese brands continue to challenge Tesla’s position in some markets, but the U.S. automaker’s rebound has been widespread in Northern and Western Europe.
Germany, Europe’s largest auto market, contributed to the positive momentum. Although full April registration data had not yet been released as of early May, March’s figures were record-setting: 9,252 Tesla vehicles registered, a staggering 315% increase year-over-year and the company’s strongest March performance in years.
Germany reported 3,149 Tesla sales and 1.3% market share in April. BEV penetration is 25.8% and Tesla has 4.9% of this segment. 🇩🇪
• +256% vs. April last year and +142% compared to January the first month of the previous quarter
• Best April ever
• Highest first month of the… pic.twitter.com/n4MIJv4w6t— Roland Pircher (@piloly) May 7, 2026
That month alone accounted for 72% of Tesla’s Q1 total in Germany (12,829 units, up 160%). Industry observers expect April to follow suit, supported by new EV subsidies and rising fuel prices.
The April figures come after Tesla’s Q1 2026 global deliveries of 358,023 vehicles, which showed modest growth but trailed some analyst expectations. The European and Chinese rebounds suggest accelerating demand heading into Q2, driven by refreshed lineups, competitive pricing, and expanding charging infrastructure.
However, Tesla faces ongoing pressure from lower-cost Chinese competitors and softening demand in select markets like Norway and Portugal, where April registrations fell sharply.
Overall, April’s data paints an optimistic picture for Tesla. The company’s ability to post consistent growth in China while reclaiming share in Europe signals renewed strength after 2025’s challenges.
Investors and analysts will watch closely for May and June numbers as Tesla prepares its Q2 report, which could confirm whether this rebound translates into sustained record-setting momentum. With approximately 450 words, this snapshot highlights how targeted execution is paying dividends in Tesla’s most critical regions
Tesla crushes NHTSA’s brand-new ADAS safety tests – first vehicle to ever pass
Tesla to fix 219k vehicles in recall with simple software update
