Stock Market News

Self-driving cars from GM’s Cruise could soon carry paying riders in San Francisco

2022.06.02 14:22

2/2
Self-driving cars from GM's Cruise could soon carry paying riders in San Francisco
FILE PHOTO: A Cruise self-driving car, which is owned by General Motors Corp, is seen outside the company’s headquarters in San Francisco where it does most of its testing, in California, U.S., September 26, 2018. REUTERS/Heather Somerville

2/2

By Paresh Dave

OAKLAND, Calif. (Reuters) – General Motors Co (NYSE:GM)’s Cruise on Thursday could become the first company to secure a permit to charge for self-driving car rides in San Francisco, if it can overcome objections by city officials.

Self-driving test cars with human safety drivers have become a constant sight in San Francisco, and completely driverless ones are increasingly common too. Turning the cars into a fledgling business in a major U.S. city would mark a milestone in the long, delayed journey toward driverless taxi service. The permit is Cruise’s final hurdle in California.

Disability and business groups have expressed support. Staff for the state commission have recommended approving Cruise, saying its proposal reasonably protects passenger safety.

Citing safety concerns, San Francisco fire, police and transit officials want state regulators to impose restrictions before allowing Cruise’s autonomous vehicles (AV) into the ride-hailing business. They recommend requiring further approval to add more cars and a new working group of state and local officials.

“Deployment of driverless AVs on a much larger scale would increase the likelihood that unusual AV behavior could lead to serious injury or death,” city officials wrote to the California Public Utilities Commission, which is scheduled to weigh Cruise’s permit application on Thursday.

The local officials say a confused Cruise AV briefly blocked a San Francisco fire engine in April that was en route to a three-alarm fire, and days earlier a driverless Cruise car stopped by police appeared to drive away before the officer was done. Cruise said its cars made safe decisions.

While rival Alphabet (NASDAQ:GOOGL) Inc’s Waymo has charged for rides in suburban Phoenix since 2018, Cruise’s proposed deployment in its hometown of San Francisco, a more densely populated, hilly and unpredictable area, is considered by tech experts the greater challenge.

Waymo has given employees driverless rides in San Francisco since March, and Cruise has offered free late-night test rides to the public since February. Cruise wants to begin by charging during the same period, in a geographic area that avoids downtown.

But there also is a longstanding issue that self-driving cars cannot always correctly predict how humans will react to changing events, including the actions of the car. Cruise has even given the issue a name, the “couples problem,” a former employee said.

ACCIDENTS AND NEAR MISSES

A decade since California first allowed public testing of self-driving vehicles, smooth rides that follow traffic rules are the norm but surprises persist.

In a public presentation last year, Cruise senior director Brandon Basso described “kinematic uncertainty,” a challenge faced by self-driving cars in predicting actions of humans on the road and deciding, for instance, when to yield.

Cruise said its vehicles understand complex social dynamics and hedge against uncertainty by taking safe actions.

Even the San Francisco officials challenging the permit said that despite “the conspicuous exceptions, the driverless Cruise AV appears to generally operate as a cautious and compliant defensive driver.”

Though self-driving cars can adapt to nearby rule-breakers, “human error or behavior such as a road rules violation that diverges from the probable behavior patterns is a factor in a disproportionate number of collisions,” Waymo told Reuters in a statement.

Cruise does not reveal what three former employees say are two key safety statistics internally: how often its cars encounter novel situations or experience what it calls “safety-critical incidents,” a combination of accidents and near-misses.

Public records seen by Reuters show Cruise with its computers in control suffered 34 accidents involving bodily harm or over $1,000 in damage across nearly 3 million miles of driving during a four-year span ended May 2021.

The documents, which Cruise unredacted in February in response to a request by Reuters, show its attempts to avoid repeat collisions.

For 28 of the cases, Cruise pursued technological fixes, which often were related to improving predictions of what humans will do. It also has relaxed some rules: including a response to a 2019 accident that allows the car to “adjust strict adherence to all marked lanes,” so that it could move around parked trucks or slow cyclists.

Waymo in January sought a court order to preserve the confidentiality of its comparable data, calling it trade secrets. The state did not oppose the request, and Waymo’s records remain redacted.

Some accidents have led to lawsuits. A bicycle courier and a scooter rider have sued Cruise, and Waymo settled over a 2016 car accident.

Cruise said the car was not self driving during the scooter incident and is fighting the case.

Bicyclist Christopher McCleary, who last month settled his lawsuit, said he has suffered ongoing injuries from crashing into a Cruise car that he said came to an unexpected halt in San Francisco in 2018, and he questions the experimentation of driverless cars in public.

“Unfortunately,” he said by email, “I feel like Cruise ‘learned’ from hitting me and this is actually a sacrifice I had to make to allow Cruise to become ‘better’ at predicting situations.”

Source

Related Articles

Leave a Reply

Back to top button
bitcoin
Bitcoin (BTC) $ 72,228.35 0.10%
ethereum
Ethereum (ETH) $ 2,652.25 1.26%
tether
Tether (USDT) $ 0.999848 0.07%
bnb
BNB (BNB) $ 596.56 1.05%
solana
Solana (SOL) $ 174.33 2.29%
usd-coin
USDC (USDC) $ 0.999891 0.02%
xrp
XRP (XRP) $ 0.52354 0.21%
staked-ether
Lido Staked Ether (STETH) $ 2,650.77 1.26%
dogecoin
Dogecoin (DOGE) $ 0.168889 2.64%
tron
TRON (TRX) $ 0.169163 2.16%
cardano
Cardano (ADA) $ 0.357426 1.29%
the-open-network
Toncoin (TON) $ 4.99 0.77%
wrapped-steth
Wrapped stETH (WSTETH) $ 3,143.68 1.49%
shiba-inu
Shiba Inu (SHIB) $ 0.000018 1.59%
avalanche-2
Avalanche (AVAX) $ 26.29 1.28%
wrapped-bitcoin
Wrapped Bitcoin (WBTC) $ 72,187.34 0.05%
weth
WETH (WETH) $ 2,658.00 1.44%
chainlink
Chainlink (LINK) $ 12.22 5.44%
bitcoin-cash
Bitcoin Cash (BCH) $ 369.83 2.68%
polkadot
Polkadot (DOT) $ 4.17 0.12%
sui
Sui (SUI) $ 2.04 0.57%
leo-token
LEO Token (LEO) $ 6.12 0.20%
usds
USDS (USDS) $ 1.00 0.98%
litecoin
Litecoin (LTC) $ 72.02 1.86%
near
NEAR Protocol (NEAR) $ 4.31 1.54%
aptos
Aptos (APT) $ 9.73 2.35%
uniswap
Uniswap (UNI) $ 8.25 3.58%
wrapped-eeth
Wrapped eETH (WEETH) $ 2,787.33 1.23%
pepe
Pepe (PEPE) $ 0.00001 1.82%
internet-computer
Internet Computer (ICP) $ 8.10 0.13%
bittensor
Bittensor (TAO) $ 499.04 3.78%
fetch-ai
Artificial Superintelligence Alliance (FET) $ 1.29 2.62%
dai
Dai (DAI) $ 0.999272 0.03%
monero
Monero (XMR) $ 161.72 1.70%
ethereum-classic
Ethereum Classic (ETC) $ 19.51 1.07%
stellar
Stellar (XLM) $ 0.094822 1.69%
kaspa
Kaspa (KAS) $ 0.110931 0.82%
ethena-usde
Ethena USDe (USDE) $ 1.00 0.04%
whitebit
WhiteBIT Coin (WBT) $ 18.81 0.96%
blockstack
Stacks (STX) $ 1.77 3.75%
dogwifcoin
dogwifhat (WIF) $ 2.58 0.17%
polygon-ecosystem-token
POL (ex-MATIC) (POL) $ 0.33143 0.15%
first-digital-usd
First Digital USD (FDUSD) $ 0.99504 0.21%
okb
OKB (OKB) $ 39.45 0.05%
aave
Aave (AAVE) $ 155.82 4.14%
immutable-x
Immutable (IMX) $ 1.38 1.69%
filecoin
Filecoin (FIL) $ 3.73 0.91%
arbitrum
Arbitrum (ARB) $ 0.556723 2.58%
optimism
Optimism (OP) $ 1.72 1.98%
crypto-com-chain
Cronos (CRO) $ 0.07558 1.37%