Automatic estimation and evaluation of multi-objective human preferences for learning from demonstration
Peer-Reviewed Publication
This month, we’re focusing on artificial intelligence (AI), a topic that continues to capture attention everywhere. Here, you’ll find the latest research news, insights, and discoveries shaping how AI is being developed and used across the world.
Updates every hour. Last Updated: 13-May-2026 22:15 ET (14-May-2026 02:15 GMT/UTC)
Researchers have explored human preferences for robot motion on a variety of household tasks. The study aimed to investigate whether preferences were similar between tasks, users, and if robots should behave in a human-like manner. The results found that preferences should be highly individualized, presenting a challenging future for integrating robots into everyday lives.
A machine learning model that analyzes patient demographics, electronic health record data, and routine blood test results predicted a patient’s risk of hepatocellular carcinoma (HCC), the most common type of liver cancer, with high accuracy.
With an expected resurgence in lunar missions in the coming years, Purdue University engineer Carolin Frueh is researching how to minimize the impact of space debris as it increases in the cislunar region.
A research team led by Jae-Pil Heo, Professor in the Department of Software at Sungkyunkwan University(SKKU), has developed an Artificial Intelligence (AI) technology that can accurately recognize new actions from only a small number of example videos.
Planting trees is widely championed as a straightforward, nature-based fix for global warming. The logic seems foolproof: expanding forests should pull more carbon dioxide from the air and pack it safely into the earth. However, a sweeping five-decade analysis of land transformation in Kerala, India, suggests the reality beneath the surface is full of unexpected trade-offs.
Published in the journal Carbon Research, the study was spearheaded by corresponding author V. K. Dadhwal at the School of Natural Sciences & Engineering, National Institute of Advanced Studies in Bengaluru. His team utilized advanced machine learning to map how half a century of plantation expansion actually impacted the dirt itself. Their findings challenge a popular assumption, proving that massive afforestation campaigns do not automatically equal a massive boost in soil organic carbon (SOC).
To accurately track the landscape from 1972 to 2020, the research team moved beyond traditional area-based counting. They fed a Random Forest predictive model with detailed historical land use maps, legacy soil measurements, local climate data, and topographic variables. This high-resolution approach allowed them to pinpoint specific geographical hotspots where carbon was either successfully sequestered or silently lost.
Arsenic contamination in rice paddies is a stubborn and dangerous threat to global food safety. Heavy metals linger in the mud, stressing the crops and eventually making their way into the human diet. While engineers frequently test expensive chemical treatments to clean up these sites, a fresh ecological approach looks to a surprisingly common material for the cure: discarded pork bones.
A newly published paper in Carbon Research explores exactly what happens when agricultural lands are treated with micro- and nano-scale bone char (MNBC). Driven by corresponding author Chuanxin Ma at the Guangdong University of Technology, the investigation proves that adding just a small amount of this specially processed biochar triggers a massive biological revival in toxic soil.
This initiative draws on the deep ecological expertise housed at the Guangdong Basic Research Center of Excellence for Ecological Security and Green Development and the Guangdong Provincial Key Laboratory of Water Quality Improvement and Ecological Restoration for Watersheds. Rather than just trapping the arsenic, the researchers discovered that the bone char fundamentally alters how the soil microbiome behaves and survives under stress.
Human–machine interface (HMI) systems require energy harvesters that can operate efficiently under low contact forces, yet conventional tactile triboelectric nanogenerators (TENGs) suffer from low surface charge density and unstable output. Here, we propose a human skin electric field-induced air-breakdown TENG (AB-TENG) with a transistor-inspired architecture. The device employs a base terminal to collect electrons from human skin via an ionized air channel formed by air breakdown, enabling efficient conversion of the skin’s electric field through two operational modes: indirect (accumulated output) and direct (instant high output). In direct mode, the AB-TENG delivers 165 V at 2 N and 290 V at 24 N, with a peak power of 22 mW—22 times higher than conventional tactile TENGs. Practical utility is demonstrated through a self-powered infrared remote control and an ultrathin keyboard. This work establishes a new design paradigm that transforms air breakdown from a limitation into a functional mechanism, advancing skin-electricity-enhanced thin-film TENGs toward next-generation self-sustaining HMI platforms.