|
|
Shell AI Effort Shows Early Returns; Health-Care AI Needs Human Touch; What Slows Driverless Car Services
|
|
|
|
|
|
|
Engineers at Shell Pernis, a refinery in the Netherlands, inspect a valve. The company has deployed predictive maintenance algorithms here and at 22 other locations world-wide, including refineries and oil rigs. PHOTO: ROYAL DUTCH SHELL PLC
|
|
|
Shell’s companywide AI effort shows early returns. Royal Dutch Shell's two-year-old strategy to embed artificial intelligence across its operations is helping the oil giant lower costs and avoid downtime, reports WSJ Pro’s Jared Council. The Anglo-Dutch company's 280 AI projects include initiatives aiming to flag equipment likely to malfunction, determine where to drill, and steer drill bits through shale deposits.
For some of its AI application development, Shell is using a software platform from software company C3.ai Inc. That platform, which uses Microsoft Corp.’s Azure cloud service, enables Shell’s AI team to build and operate AI and internet-of-things applications at scale.
Keeping operations running. One of Shell’s main AI projects focuses on predictive maintenance, spotting potential faults with valves, compressors and other extraction or production equipment. Sensors, both wired and wireless, capture equipment-performance data. The sensors feed machine-learning algorithms trained on historical data, including conditions such as the temperature and pressure of internal parts, during past malfunctions. By learning which conditions preceded mechanical “trips,” those algorithms have provided early warnings that allowed employees to replace parts before malfunctions could halt operations.
Helping to locate new oil and gas sources. Another large AI project is aimed at helping the company find new sources of oil and gas by cleaning up data from seismic surveys, which are used to create images of rock formations that in turn help scientists locate oil deposits below the ocean floor. The problem, historically, has been that these surveys don’t paint a clear picture of what rock formations look like. Underwater currents and other factors produce noisy data that affects the images. Shell created machine-learning algorithms, based on images the company has cleaned, to filter out that noise. Those surveys used to take human workers several months to interpret. The AI system reduced the time needed to produce clearer images by 80%.
|
|
|
|
|
A doctor using a digital tablet in a hospital. PHOTO: CREDIT: GETTY IMAGES
|
|
|
Medicine without doctors doesn’t compute. Health-care providers are finding many uses for AI. For instance, the technology is being used to detect eye disease and interpret medical images. In addition, some providers have started to use AI-based software to ask preliminary questions about symptoms that a caregiver would normally ask, reports the Associated Press. However, while AI and other innovations are creating health-care applications that may seem like science fiction, Dr. Marc Siegel, a clinical professor of medicine and medical director of Doctor Radio at NYU Langone Health, writing for the WSJ, says “none of them will fly without the literal hand-holding that only a flesh-and-blood physician can provide. Patients won’t accept some of the more dystopian innovations on the horizons of personalized care—retina scans, nanomedicine, bioengineered immunities—unless their treatment is being driven by actual doctors who learned their jobs the old-fashioned way.”
|
|
|
|
|
The challenge of commercialization. The technical development of autonomous cars looks to be nearing completion, with the next logical step being mass deployment, reports the Financial Times. However, there is increasing acknowledgment that the commercialization of driverless autos is a big challenge “that requires government approval, public trust, brand marketing, the ability to manufacture at scale and the technical know-how to manage a fleet that competes with the likes of Uber and Lyft on timely pickups,” according to the report. Case in point: General Motors’ driverless-car business, Cruise, over the summer delayed the roll out of an autonomous ride-hailing service, which the Journal reported underscored the difficulty of safely deploying robotic vehicles on public roads.
|
|
|
|
Startup banks on sanctions compliance. Singapore startup Tookitaki Holding Pte. Ltd. raised $11.7 million to sell its financial fraud detection software to banks as regulators push the sector to comply with U.S. sanctions, reports WSJ Pro’s Marc Vartabedian. Tookitaki’s software helps financial institutions spot nefarious behavior related to money laundering. Artificial intelligence in the software can automatically detect suspicious patterns in money movements and the use of shell companies, among other things, said Abhishek Chatterjee, Tookitaki’ co-founder and chief executive.
|
|
|
|
Uber lost its license to operate in London after regulators said they found instances of unauthorized drivers swapping their own photos with those of authorized drivers on Uber’s platform. The regulator said Uber addressed the problem but was worried about drivers finding ways around other restrictions. Uber said it would appeal and that it would soon introduce “facial matching” to prevent what it called photo fraud. (WSJ)
Delta Air Lines is implementing facial recognition to confirm the identities of international travelers at Seattle-Tacoma International Airport. (Seattle Times)
Taiwanese startup Appier raised $80 million to advance its AI-based marketing and customer-engagement software. (TechCrunch)
Hyundai is planning to test driverless cars in Seoul. (TechCrunch)
Piaggio, which makes the Vespa scooter, is marketing a cargo-carrying robot. (Associated Press)
LG has installed a robot that will make and serve noodles at a South Korean restaurant. (ZDNet)
Amazon unveiled a new tool that can help customers train machine learning models on limited data sets. (TechCrunch)
|
|
|
|