Connect with us

News

Mobileye CTO on Building Autonomous Driving Systems

Published

on

Autonomous driving system

Mobileye 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.

Mobileye autonomous driving

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

"I write about technology and the coming zero emissions revolution."

Advertisement
Comments

News

Tesla Cybercab uses a unique strategy for picking up the right rider

Published

on

Credit: ARTSIMAGE | X

Tesla Cybercab is using a unique strategy for picking up the correct rider, which is a crucial part of ride-hailing to ensure people end up in the right place and are charged the correct price.

Cybercab will utilize an RGB strip in its front light bar that will illuminate in a variety of different colors to mark itself.

This identifying mark will also appear in the Robotaxi app, giving riders in the same location a notable distinction in an effort to avoid any confusion regarding who should get in each vehicle.

Other ride-hailing services use similar strategies: Lyft and Uber rides are recognizable through driver identity, vehicle type and color, as well as license plate. Waymo will display the rider’s initials on top of the vehicle, letting them know that the specific vehicle for them has arrived.

Tesla’s strategy is unique and interesting, but there are some flaws. Cybercab’s main purpose is aimed toward being an autonomous ride for all, including those who have disabilities like being blind or even color blind.

Tesla will likely have something in the pipeline for those who cannot see colors or have limited vision. There will definitely be multiple ways to identify which vehicle is the one that “you” specifically ordered.

Cybercab is set to start giving public rides next Thursday, September 3, in Austin, as it announced a dedicated event last week and invited many members of the Tesla community.

Advertisement
-

Tesla will launch Cybercab on September 3

Additionally, members of the public will be invited as well. Tesla has been offering employee rides in Cybercab for nearly two months.

Continue Reading

News

Tesla ends in-house wrap service that always seemed like a short-term program

Published

on

Credit: Tesla

Tesla has said goodbye to one customization option for its vehicles: the wrap service it launched several years ago.

After launching an in-house wrap service in August 2020 for the first time in China. In the U.S., it launched in October 2023. Tesla continued to expand the program and adjust it with better pricing and fewer options for the Cybertruck.

By December 2023, it was giving owners of the Model 3, Model Y, and the Cybertruck the opportunity to give their vehicle a fresh look with a vinyl wrap.

Tesla revamps in-house vinyl wrap service with better pricing

It was only available in five locations: Costa Mesa, Oceanside, Santa Clara, West Covina, all in California, and Seattle, Washington.

However, Tesla made some big adjustments to its shop, and the wrap service is officially gone:

Advertisement
-

Wraps are very popular across the Tesla lineup, especially since the company offers relatively few colors. Many choose to wrap their Teslas with interesting colors, patterns, or even finishes, turning their cars from glossy to satin or matte.

However, Tesla’s wrap service was so limited geographically that it never really had a chance to get off the ground or compete with local shops. Every area in the United States is now overflowing with detailing shops, mobile detailers, and other automotive specialists, many of whom perform wrap services.

Tesla’s service was confined to the Pacific time zone and only spanned across two states. It was never going to be something Tesla was a major competitor in, nor was it going to disrupt the wrapping industry. Now that the program has ended, it seems pretty ideal to believe it was always going to be a short-term thing.

Along with the wrap service, Tesla removed several other products, but nothing too crazy. The Model 3 Door Pocket and Cupholder Liners, the Model S 19″ Magnetite Wheel and Winter Tire Package, Model X/Y Ski/Snowboard Carrier for Hitch Rack, Tesla’s Electric Summer Party Tee, and the Electric Summer Tee were the other items the company totally eliminated from its online shop.

Continue Reading

News

Tesla Robotaxi fleet gets a brain upgrade ahead of Cybercab launch event

Tesla’s Robotaxi service now runs longer hours nationwide as its unsupervised fleet quietly grows larger.

Published

on

By

Tesla’s Robotaxi service just got easier to catch, with the company’s official Robotaxi account noting that rides are now available from 6 a.m. to 10 p.m., seven days a week, across its operating footprint. The account also said its unsupervised fleet is “a lot bigger” than before, though without specifics. The bigger change is what Tesla says upgraded intelligence in vehicle distribution and routing is what’s actually cutting wait times, not a new Full Self-Driving version.

While Tesla did not name the team behind the upgrade, the language points to its AI and fleet software group rather than the driving stack itself. Vehicle distribution and routing in Robotaxi has functioned mostly as a dispatch problem with the software deciding which idle car goes to which rider, and how far it has to travel to get there. “Upgraded intelligence” suggests a smarter version of that dispatch logic, likely using demand forecasting to position idle cars near where riders are about to request them rather than reacting once a request comes in. Tesla’s AI division has built similar prediction systems for other parts of the business, including the neural networks that power FSD itself, so applying that same approach to fleet logistics would be a natural extension rather than a new discipline for the team.

Tesla is also about a week away from a separate robotaxi milestone. The company plans to launch Cybercab, its purpose built two seat robotaxi with no steering wheel or pedals, in Austin on September 3. Cybercab has been giving employees rides on public and private roads for weeks, and the September event is expected to fold those vehicles into the existing Robotaxi fleet within days of the launch.


Austin previously ran Robotaxi from 6 a.m. to 2 a.m. as of last September, a schedule set before the service expanded into Dallas, Houston, Miami, Tampa, Orlando and the Bay Area. Wednesday’s post did not specify whether that extended overnight window still applies in Austin specifically or whether 6 a.m. to 10 p.m. is now the standard across every market. Tesla’s post, visible on its official Robotaxi account, framed the change simply as fewer riders waiting around for a car.

Whether the wider hours hold once Cybercab enters the fleet next week is the next thing worth watching. Tesla has tended to expand Robotaxi in increments, first geofence, then hours, then fleet size, and each step so far has arrived without much advance notice.

Advertisement
-
Continue Reading