TRANSPORTATION AND LOGISTICS BRIEFING: GM's first self-driving car model - New regulatory guidelines for self-driving cars - Thermal sensors for fully driverless vehicles
Welcome to Transportation & Logistics Briefing, a new morning email providing the latest news, data, and insight on how digital technology is disrupting transportation and delivery, produced by BI Intelligence.
Have feedback? We'd like to hear from you. Write me at: jcamhi@businessinsider.com.
GM AND CRUISE DESIGN THEIR FIRST SELF-DRIVING CAR FOR MASS PRODUCTION: GM and Cruise, an autonomous car startup that GM acquired for more than $1 billion last year, announced that they have finished the product design for their first mass-produced self-driving vehicle, Tech Crunch reports. That means the auto giant has all of the hardware components in place to launch a commercial self-driving vehicle.
The self-driving car model, dubbed the Cruise Generation 3, is based on the design of the electric Chevy Bolt, and will be produced in GM's assembly plant in Orion, Michigan. GM said that it can produce tens of thousands of Generation 3s per year, once the vehicle is ready for commercial launch. The Generation 3's design includes redundant systems, so if one sensor or computer fails, another will come online to ensure that the car remains safe. For now, the Generation 3 includes a steering wheel and pedals for a human driver, but GM suggested that it plans to eventually remove those. The automaker had to modify about 40% of the Bolt's components to retrofit the Generation 3 for self-driving, but it will retain the Bolt's range of 238 miles on a single charge.
GM plans to manufacture fleets of Generation 3s for services like ride-hailing and package delivery. All of the additional technology in the Generation 3 will give it a hefty price tag out of the reach of most consumers. GM said it will test 50 new Generation 3s with its employees over the next few weeks, and drivers will be present for all of the tests. Those 50 vehicles will be integrated into the Cruise Anywhere ride-hailing service, which offers on-demand rides to Cruise's employees in San Francisco. GM could both launch its own ride-hailing service using Generation 3s and sell fleets of Generation 3s to other ride-hailing companies. The company has a major partnership with Lyft centered around developing self-driving vehicles for Lyft's ride-hailing app, and GM's Maven car-sharing division has hinted that it may offer on-demand rides in the future as well.
Only two roadblocks remain on the path to the Generation 3's commercial launch - further development of self-driving software and regulatory approval. GM said that the vehicle is designed without the need for a driver, but self-driving software needs further testing and improvement before self-driving cars are capable of handling any scenario they may encounter on the road. GM purchased Cruise specifically to use its software kit to accelerate GM's development of self-driving systems. It will likely take years though for Cruise and GM to develop self-driving software that can fully replace a driver. Likewise, regulatory approval for self-driving cars without a driver is still far off, as the National Highway Traffic Safety Administration (NHTSA) faces shortages in staffing and resources for overseeing self-driving cars. However, legislation recently passed by the House of Representatives would provide expanded exemptions for automakers to test more self-driving vehicles, allowing them to vastly expand the scope of their self-driving projects.
BII
US DEPARTMENT OF TRANSPORTATION RELEASES NEW SELF-DRIVING GUIDELINES: The US Department of Transportation (DoT) released an updated version of its non-binding self-driving car guidelines on Tuesday, The Washington Post reports.
The new version cut some guidance from the previous iteration that the department released last year in an effort to streamline the development of self-driving cars. The DoT trimmed the previously 15-point safety assessment for self-driving cars down to 12 points, eliminating sections on privacy, vehicle registration, and ethical considerations. The assessment still includes guidance on crash-testing, recording of crash data, cybersecurity, object detection and avoidance, and consumer education.
The new version also asserts that regulatory authority over new car technologies - including self-driving systems - lies with the federal government instead of state agencies. This is a major win for auto and tech companies developing self-driving technology. In the absence of federal regulations, 20 US states have passed their own rules for self-driving cars that overlap and sometimes conflict with each other. The DoT's guidelines though clearly state that federal rules will take precedence in regards to setting and enforcing safety standards, removing the burden of complying with all of these various state regulations. However, states will retain their existing authority to formulate and enforce rules around vehicle registrations and inspections, as well as insurance and liability.
While automakers responded favorably to the new guidelines, consumer advocacy groups expressed concerns that the DoT isn't taking a more proactive approach to self-driving regulation. The 12-point safety assessment is completely voluntary for companies that are developing self-driving cars - the guidelines are meant to provide a blueprint for what factors companies should consider when developing self-driving technology, but are not legally enforceable. However, a bill recently passed by the House of Representatives would make the guidelines legally enforceable if it's approved by the Senate and passes into law.
Enjoy reading this briefing? Sign up and receive Transportation & Logistics Briefing to your inbox.
THERMAL IMAGING'S SELF-DRIVING POTENTIAL: AdaSky, an Israeli automotive startup, unveiled a new far infrared thermal camera, known as Viper, that's designed to work with computer vision algorithms to help self-driving cars detect objects around them. The company didn't specify when the camera would be available, or for how much.
AdaSky is betting Viper can help self-driving cars handle scenarios that other underlying technologies can't. Most self-driving cars rely on radar, cameras, and LiDAR - a laser-image sensing technology - to help create a virtual map of the world around them. Viper isn't designed to replace these technologies, instead the company wants Viper to work with them, and help the cars view the roads when these other technologies can't do the job. That could include scenarios in inclement weather or at night involving oncoming headlights, or sudden flashes of light at dawn or dusk that blind the car's cameras and sensors temporarily. All of these situations are often difficult for self-driving vehicles to handle.
That means there's an opportunity for AdaSky with Viper, but it's a long-term one. Right now, most companies developing self-driving cars are looking to place only semi-autonomous cars on the road that still have a human in the driver's seat to take control of the car in situations where the car's systems fail. But in the long-run, these companies are aiming to develop fully autonomous cars that don't require a driver at all. That means they'll need technologies to help them handle every possible scenario, which could prompt companies like Intel or Waymo to build new camera equipment, like Viper, into their self-driving systems.
In other news…
- China's government will ban the sale of fossil fuel vehicles entirely, though it hasn't set a date yet for the ban to take effect, according to Bloomberg. The country's regulators are currently working with automakers to develop a timeframe for when the production and sale of the vehicles will be banned. The move is part of the country's larger effort to limit carbon emissions by 2030 to reduce air pollution. This move will also likely boost self-driving car development in the country, since the systems that power self-driving cars work better with electric vehicles than with fossil fuel powered ones.
- Mapping and navigation giant TomTom will supply its maps to German auto giant Daimler for the automaker's North American cars. The maps can be used by drivers and passengers through the in-car infotainment centers of new Mercedes A, B, C, and E-class sedans. The move could boost the in-car experience of these Mercedes vehicles, since TomTom offers some of the most advanced mapping and navigational systems currently available. Further, it's possible TomTom could also supply maps for the self-driving cars and semi-trucks that Daimler is working on, since high-quality maps are a necessity for autonomous vehicles.
- XPO Logistics, a Connecticut-based logistics service provider that specializes in transporting heavy goods, is expanding its network for last-mile deliveries with 85 new hubs across the US. XPO operates a massive long-haul trucking and rail freight network, but the expansion will help the company boost its home delivery and e-commerce fulfillment offerings. The new hubs will be located in major urban centers such as Los Angeles, Buffalo, Milwaukee, and Washington D.C. The company plans to have the new hubs up and running by the end of next year, at which point its last-mile network will reach 90% of the US population. The move should help the company capitalize on the rapid growth of e-commerce.