Car Hospital

20/09/2026

https://www.facebook.com/share/p/1DWc4gVLpR/

Bathurst is not a circuit where foreign cars are supposed to win. Mount Panorama is 6.2 kilometers of public road on a mountain, it rains without warning, and Australian touring car racing was built on rear-drive V8s.

In 1991, Jim Richards and Mark Skaife won the Bathurst 1000 in a Gibson Motorsport R32 GT-R by a margin the local press struggled to describe. The car was Japanese, the engine displaced half what the Commodores and Falcons carried, it had all-wheel drive, and it had two turbochargers. The word the media settled on was Godzilla, and it stuck permanently.

The engine explains the result.

Two point six liters. Bore 86 mm, stroke 73.7 mm, 2,568 cc, which is 156 cubic inches. DOHC, 24 valves. Two ceramic Garrett T25 turbochargers, one feeding three cylinders each.

That arrangement matters. A single large turbocharger sized for peak power takes time to spool, and the engine feels dead below the threshold. Two smaller turbos each handling half the exhaust flow spin up far more quickly, so boost arrives earlier and the engine responds when the driver asks rather than a moment later.

Then six individual throttle bodies, one per cylinder. A single throttle body sits ahead of a plenum, and every throttle input has to travel through that shared volume before the cylinders feel it. Individual throttles sit directly on each runner, so the response is immediate and every cylinder gets the same mixture through the same runner length.

The road car was rated at 276 horsepower, which was not a real number. Japanese manufacturers had a gentleman's agreement capping published output at exactly 276 regardless of what the engine made, and the RB26 was closer to 320.

In Group A race trim, Gibson Motorsport ran it at roughly 550 to 600 horsepower with more boost, race mapping, and upgraded internals.

The other half of the package was ATTESA E-TS, which distributed torque between the front and rear axles electronically rather than through a fixed mechanical split. On a wet mountain circuit with severe elevation changes, a car that can send power wherever the grip is has an advantage no rear-drive car can answer.

The record is what makes it remarkable rather than just fast.

In the Japanese Touring Car Championship, the R32 won all 29 races it entered between 1990 and 1993. Every competing manufacturer withdrew and the series collapsed.

In Australia it took the touring car championship in 1991 and 1992, with Richards and Skaife finishing first and second both years. Their 1991 Bathurst race time stood as a record for nineteen years.

At the 1991 Spa 24 Hours it won outright and set the fastest lap, beating European teams on a European circuit.

Three continents, at the same time, from an engine displacing 156 cubic inches.

19/09/2026

Να! Αυτά με τρελαίνουν, που να βρεις τέτοια μαστόρια...

https://www.facebook.com/share/v/1FWWi93vEB/

14/08/2026

11/08/2026

https://www.facebook.com/watch/?v=27233273926351449

01/07/2026

🤣🤣🤣🤣

Ford has revealed that it brought back hundreds of experienced engineers after artificial intelligence failed to match the expertise of human inspectors during vehicle quality checks. The decision highlights an important lesson for manufacturers: while AI can improve efficiency, it still depends heavily on the knowledge and experience of skilled people.

Like many major companies, Ford invested heavily in AI to streamline its operations. The company introduced AI-powered systems across different stages of manufacturing, including quality inspections. Around 900 AI-enabled cameras were installed in factories to detect defects early, improve production standards, and reduce costly disruptions. The goal was to make inspections faster, more consistent, and more efficient.

However, company executives admitted that the technology did not perform as well as expected. According to Charles P**n, Ford's Vice President of Vehicle Hardware Engineering, the company underestimated the value of its most experienced engineers. He explained that AI is only as effective as the information used to train it. Without decades of practical knowledge from veteran engineers, the automated systems struggled to identify many of the subtle quality issues that experienced inspectors could easily recognize.

Over previous years, many highly skilled engineers had left the company before their expertise could be fully incorporated into Ford's AI systems. As a result, the automated tools lacked the real-world judgment developed through years of designing, testing, and inspecting vehicles across multiple product generations.

To address the problem, Ford rehired more than 300 veteran quality engineers and inspectors. Their role extends beyond simply inspecting vehicles. They are now helping train AI systems with real manufacturing knowledge while also mentoring younger engineers entering the workforce. By combining human expertise with machine learning, Ford hopes to create more reliable inspection systems in the future.

P**n acknowledged that the company had mistakenly believed feeding design specifications into AI would automatically produce high-quality results. Instead, Ford learned that manufacturing quality depends on much more than technical data. Experienced engineers often recognize patterns, unusual defects, and production issues that are difficult to capture using software alone.

Ford's renewed focus on experienced personnel appears to have delivered positive results. The company recently regained the top position among mainstream automakers in the J.D. Power Initial Quality Study, an industry benchmark measuring vehicle quality. It marked the first time Ford had achieved this ranking since 2010.

In announcing the achievement, Ford credited a significant overhaul of its engineering and manufacturing teams. Alongside leadership changes across engineering, supply chain, and production, the company emphasized that bringing back veteran engineers played a major role in improving vehicle quality. Their decades of hands-on experience helped strengthen both manufacturing processes and AI training.

Ford's experience reflects a broader reality facing many industries adopting artificial intelligence. While AI can analyze enormous amounts of data, automate repetitive tasks, and improve productivity, it cannot instantly replace years of practical human experience. Instead, the most effective approach often combines advanced technology with skilled professionals who understand the complexities of real-world decision-making.

The company's decision serves as a reminder that AI is a powerful tool—not a complete replacement for human expertise. In industries where precision, safety, and quality are essential, experienced engineers continue to provide knowledge and judgment that even the most advanced AI systems are still learning to match.
#

06/05/2026

https://fb.watch/GWUQifGW-w/

28/04/2026

Ετοιμαστείτε, καλή η τεχνολογία ε;
Καλή, αλλά παίζει ρόλο ποιος την ελέγχει και με τι σκοπό την διαχειρίζεται!

🚨 Federal Law Forces Invasive Driver Surveillance Systems into Every New U.S. Car by 2027

Congress enacted Section 24220 of the 2021 Infrastructure Investment and Jobs Act, directing the National Highway Traffic Safety Administration to mandate advanced impaired driving prevention technology in all new passenger vehicles.

This federal requirement locks in for model year 2027, with rollout accelerating toward late 2026.

Infrared cameras, AI-powered algorithms, and biometric sensors will embed directly into vehicle dashboards nationwide.

These systems will relentlessly scan eye movements, head position, pupil changes, breathing rhythms, steering behavior, and other vital signs in real time.

Detection of suspected impairment, whether accurate or triggered by a false positive, triggers automatic intervention: the car refuses to start or abruptly limits speed, stranding drivers without appeal.

NHTSA’s own February 2026 report to Congress admits current technology falls far short of required accuracy, warning that even high detection thresholds could generate millions of erroneous shutdowns annually.

Yet the statutory mandate presses forward unchanged. Congressional efforts to repeal or defund it (including amendments backed by Rep. Thomas Massie) collapsed in January 2026, sealing the timeline.

Every new vehicle will carry an extra several hundred dollars in hidden costs.

The law creates a vast, uncharted biometric data ecosystem inside private cars, with no ironclad protections against hacking, insurer access, or warrantless government queries.

Sensitive personal information—your gaze, fatigue levels, potential medical signals—will flow through systems ripe for abuse as surveillance capabilities expand.

This marks a chilling federal takeover of personal mobility.

Americans will sit under constant algorithmic judgment every time they enter a new vehicle, with machines empowered to override human decisions on the open road.

I do not consent to this government-mandated surveillance embedded in my car.

Do you?

11/04/2026

Σας ευχόμαστε καλή ανάσταση!
Καλή επιστροφή με ασφάλεια!

25/02/2026

https://www.facebook.com/EngineeringUni/posts/pfbid0pSBP8U7kbJS5TyNgSaZ4Ysm37qsUD3iWtXmiyfZvvH85HRBQY7pGft5q482ZYTUnl

In the 1980s turbo era, Formula 1 rules were simple: turbo engines were limited to just 1.5 liters. To win, BMW had to run crazy levels of boost that would push many conventional racing blocks past their limits.

Casting brand‑new blocks sounded like the obvious fix. But freshly cast iron can keep hidden internal stresses from cooling unevenly, which can make it more likely to warp or crack under brutal heat and pressure.

BMW engineer Paul Rosche went the opposite way. Instead of new castings, his team hunted down used M10 4‑cylinder blocks from regular road cars with high mileage, commonly described as well over 100,000 km. Years of heating and cooling on the street helped “season” the iron and reveal any weak castings before they ever got near an F1 car.

Those seasoned blocks became the base for BMW’s M12/13 monsters. For Saturday qualifying, they dropped in F1‑grade internals, hung a huge K*K turbo on the side, and cranked boost to several bar, often quoted around 4 bar, roughly 60 psi, in some qualifying trims. Power figures for full qualifying trim are commonly reported as well above 1,000 horsepower.

To stop the engine from detonating itself to death, they ran a special race fuel that was heavily rich in toluene. This high‑octane, stable fuel helped resist knock and soaked up heat as it vaporized, acting almost like a chemical intercooler inside the cylinder.

These qualifying setups were known as “grenade” engines. They were built to survive only a few laps, an out lap, one flat‑out flyer, and an in‑lap, before being taken out of service. After qualifying, the team swapped them for a more durable, lower‑boost version, often reported around 800 horsepower, for Sunday’s race.

It was asymmetric engineering at its best: the heart of one of F1’s wildest engines started life as a worn‑in block from a normal road car.

For more deep-dive engineering stories like this, join the thousands of engineers reading our weekly newsletter. Link in bio.

22/02/2026

Βρε δεν το ξανακάνω σ' Autobianchi... 🤣🤣