This week’s roundup of data news highlights from September 13, 2026, to September 18, 2026, features MIT’s robotic laboratory for automating optics experiments and an AI model that responds to human expression during real-time conversations.
1. Improving Road Safety
England’s National Highways department has installed an AI-powered camera at a crash hotspot in Cornwall to detect dangerous U-turns. The camera analyzes traffic movements to identify vehicles making prohibited U-turns and flags potentially dangerous maneuvers. It monitors traffic continuously and improves safety on high-risk roads by helping authorities identify risky driving without requiring officers’ continuous attention.
2. Upgrading Warehouse Robots
Oregon-based robotics company Agility Robotics has built the latest version of its Digit humanoid robot, Digit 5, which can perform a wider range of warehouse and factory tasks. Unlike earlier models that mainly moved containers, Digit 5 can unload pallets, tend machines, gather parts, and inspect items before shipment. The robot could help companies automate more physical work across warehouses and manufacturing facilities.
3. Testing Responsible AI
Sheffield Hallam University in the UK has launched its Centre of Excellence in AI and Robotics to advance ethical, inclusive, and human-centered AI research. The center is developing projects ranging from robots that learn in ways inspired by children’s brains to autonomous systems that help stroke survivors with rehabilitation. Its research aims to ensure developers consider people who may be overlooked when creating new technologies.
4. Assessing Food-Safety Risks
Chipotle has partnered with software company Palantir to test a new food-safety risk- management platform that brings information from across its restaurants into one system. The platform combines data such as health department inspections, pest incidents, and employee illnesses to identify trends and assess food-safety risks at various locations. The system aims to help Chipotle prioritize and address potential problems before they become more serious.
5. Diagnosing Stroke Patients
Clinicians at the Department of Neurology at Korea University Guro Hospital in South Korea have tested an AI system with standard brain CT scans to detect significant blockages in brain blood vessels that can cause severe strokes. The system identifies signs of blockages without requiring contrast agents. Researchers tested the system on 900 patients in South Korea and the U.S., where it has helped clinicians identify stroke patients who might otherwise be missed.
6. Automating Optics Experiments
Researchers at MIT have built a robotic laboratory that can automatically assemble and fine-tune experiments involving lasers, mirrors, and lenses. A robotic arm identifies and positions optical components, while motorized tools adjust their angles to align beams of light. The system can also dismantle experiments and reconfigure the components for new setups. The laboratory could help scientists run complex optics experiments faster and with less manual work.
7. Streamlining Record Inspections
The Department of Veterans Affairs has integrated Microsoft’s AI Copilot into health-care inspections to help investigators review large amounts of information more efficiently. The tool analyzes interview transcripts, documents, and questionnaires using standardized prompts to identify trends and potential issues. Investigators verify AI-generated findings against original sources before including them in reports, keeping humans responsible for final conclusions.
8. Increasing Surgical Precision
Researchers at MIT have created an AI system that matches real-time X-rays taken during surgery with a patient’s 3D CT or MRI scan. The system uses the patient’s scan to generate synthetic X-rays, allowing an AI model to learn how the patient’s anatomy appears from different angles and align new X-rays within seconds. The technology could help clinicians navigate surgical tools more precisely during minimally invasive procedures.
9. Understanding Human Expressions
Seattle-based AI startup Nuance Labs has created an AI model capable of understanding and responding to human expression during real-time conversations. The model processes cues such as speech, tone, gaze, facial expressions, and timing simultaneously, allowing it to generate matching vocal and facial responses. The technology could make AI avatars more natural and responsive in applications such as customer service, education, sales, and professional coaching.
10. Lowering the Wage Gap
Researchers at the University of California, Santa Barbara partnered with the University of Toronto and Zhejiang University in China to test how AI can help disabled workers overcome workplace barriers and narrow wage gaps. The study examined deaf and hard-of-hearing food delivery workers who used an AI tool that converted text into speech, allowing them to call customers and resolve delivery issues.
