Teaching Drones to Play Tag with AI

Researchers at Sandia National Laboratories are using reinforcement learning (RL) to teach autonomous drones how to cooperate and make rapid decisions in a strategic version of tag. In the experiment, two evader drones attempt to reach a protected base while two pursuer drones try to intercept them. The evaders learn through simulated rewards and penalties to coordinate their movements, including strategies that can cause the pursuers to interfere or collide with one another. Unlike the RL-trained evaders, the pursuers use a conventional proportional-navigation algorithm. The research demonstrates how multi-agent reinforcement learning can produce flexible cooperative behaviours without engineers having to explicitly program every possible manoeuvre.

The project is overcoming the simulation-to-reality gap: strategies that perform well in simulation can fail on physical drones because of factors such as battery depletion, aerodynamic interactions, communication delays and other unmodelled dynamics. Sandia therefore tests the algorithms on small, inexpensive quadrotors in its CAMINO facility, using high-precision motion capture to evaluate and refine their behaviour before moving toward more costly systems. Beyond the game itself, the work could contribute to autonomous systems capable of pursuit, evasion and coordinated swarm behaviour in rapidly changing environments, including potential applications for protecting critical infrastructure against hostile drones and other national-security scenarios.

More information:

https://www.sandia.gov/labnews/2026/08/13/teaching-drones-to-play-tag/

Humanoid Robots Take on Real-World Firefighting

At the 2^nd World Humanoid Robot Games (WHRG) in Beijing, 23 teams competed in a realistic simulated firefighting challenge aimed at testing whether humanoid robots can progress from impressive demonstrations to practical work in hazardous environments. Held at an actual fire brigade, the competition required robots to complete three tasks within 30 minutes: identify hazardous materials, locate and close open valves, and find and operate a fire extinguisher. The event exposed significant challenges associated with real-world deployment, including rain, changing lighting, object recognition, manipulation accuracy and movement speed. Of the 12 teams competing on Sunday, only three completed the full challenge, highlighting the gap between laboratory performance and reliable operation in unpredictable environments.

Despite these limitations, organisers and participating teams viewed failures as valuable opportunities for improving humanoid robotics. Data from unsuccessful object recognition, manipulation and motion-planning attempts can be used to refine algorithms and train AI models. The competition also demonstrated emerging approaches such as VR-based teleoperation, multi-jointed robotic hands and specialised actuators combining precision with strength. The event reflects China's broader effort to move humanoid robots and embodied AI into practical applications, particularly dangerous tasks such as emergency response. Ultimately, the objective is not necessarily to replace firefighters, but to develop robots capable of working alongside humans and undertaking operations where human exposure would be risky.

More information:

https://www.globaltimes.cn/page/202608/1368322.shtml

AI Assistant for Art Provenance

Researchers are exploring how AI can help solve one of the art world’s most difficult problems: establishing an artwork’s provenance, or documented history of ownership and attribution. The NPR report describes efforts to develop an AI-based provenance assistant capable of searching and connecting information scattered across archives, auction catalogues, museum records, historical documents and other sources. Rather than replacing provenance researchers, the technology is intended to accelerate the labor-intensive process of identifying names, dates, transactions and relationships that may reveal where an artwork has been and who owned it. This could be particularly valuable for works whose histories contain significant gaps or whose ownership changed during periods such as war, displacement or political upheaval.

At the same time, the approach highlights the limitations of using AI for sensitive art-historical research. Provenance evidence can be incomplete, contradictory or ambiguous, meaning that an AI-generated connection cannot automatically be treated as historical fact. Human scholars still need to examine original records, assess the reliability of sources and determine whether proposed connections are credible. The broader promise, therefore, lies in using AI as a research assistant rather than an authority: a tool capable of navigating enormous collections of digitised information and suggesting leads that experts can subsequently verify. Such systems could ultimately make provenance investigations faster and more systematic while helping museums, collectors and researchers address questions of authenticity, ownership and potentially contested cultural property.

More information:

https://www.npr.org/2026/08/06/nx-s1-5922729/ai-art-provenance-assistant

Robots That Make Recycling Smarter

Researchers at the Karlsruhe Institute of Technology (KIT) have developed a robotic system designed to automatically dismantle broken products while preserving valuable components for reuse. Unlike conventional robotic assembly, disassembly is difficult because damaged devices can behave unpredictably. The system combines robotic manipulators with predictive algorithms, CAD models, and mathematical models of possible damage. It tests how individual components move, compares their behaviour with expectations, and updates its disassembly plan in real time. For example, if a screw is stuck, the robot can recognize that its original strategy has failed and switch to another technique, such as milling away surrounding material.

Users can also specify which components are particularly valuable, allowing the system to prioritize keeping them intact. The researchers see technology as an important step toward an automated circular economy, in which machines and electronic products are repaired, refurbished, and reused rather than discarded. The long-term vision is a factory containing many specialized robotic arms equipped with different tools, capable of dismantling a wide range of products, extracting defective components, replacing them, and potentially rebuilding the devices automatically. If scaled successfully, automated disassembly could make repairing products cheaper than manufacturing replacements, reducing electronic waste and the demand for new raw materials.

More information:

https://spectrum.ieee.org/recycling-robot

AI Creates More Realistic Multi-Person Images

Researchers at Cornell have developed a new AI approach for generating more realistic images involving multiple people interacting with one another. While existing image-generation models can create convincing individual people, they often struggle to represent complex interactions accurately. The Cornell method addresses this through iterative pose-image generation, progressively constructing a scene one person at a time. Each predicted pose helps guide the generation of subsequent people, allowing the system to better capture spatial relationships and interactions without requiring users to manually specify poses. The approach uses the FLUX image-generation model as its foundation and combines pose detection with a multimodal large language model to organise descriptions, poses and spatial regions for each person.

The researchers also introduced DrawWaldoWorlds, a benchmark designed to evaluate whether image-generation systems correctly represent not only multiple individuals but also their roles and relationships, essentially testing whether the model understands who does what to whom. Experiments showed that the new approach produced more faithful multi-person scenes than existing methods. In a user study involving 20 participants, images generated using the Cornell method were preferred roughly two-to-one over images produced by two versions of FLUX. The researchers argue that automatically incorporating pose information could make generative AI considerably better at depicting complex social activities, sports, group scenes and other situations where realistic human interaction is essential.

More information:

https://news.cornell.edu/stories/2026/07/strike-pose-creating-more-realistic-multi-person-images

Zoox Wins U.S. Approval for Driverless Robotaxis

Amazon-owned Zoox has become the first company in the United States to receive federal approval for the commercial deployment of purpose-built robotaxis that have no steering wheel, pedals or other human driving controls. The exemption from the National Highway Traffic Safety Administration (NHTSA) allows Zoox to deploy up to 2,500 vehicles annually for the next two years and begin charging passengers, initially in Las Vegas before expanding to other markets subject to state and local approvals. The electric, carriage-style vehicles feature inward-facing seats and are designed entirely around autonomous operation rather than modified versions of conventional cars.

The approval represents an important regulatory milestone for autonomous transportation, but it comes with enhanced safety oversight. NHTSA will require Zoox to provide additional reporting on crashes, inappropriate road stops and other operational problems, and the agency can withdraw the exemption if significant safety concerns emerge. Regulators are particularly focused on how autonomous vehicles interact with emergency services, construction zones, school buses and unusual road conditions following several incidents across the industry. The decision also signals a broader shift in U.S. regulation toward developing safety standards specifically designed for fully autonomous vehicles.

More information:

https://www.reuters.com/world/amazons-zoox-wins-first-us-approval-paid-robotaxis-with-no-human-controls-2026-07-30/

Digital Innovation Is Transforming the Museum Experience

Museums around the world are embracing digital technologies to create richer, more engaging experiences for visitors. By using data to better understand how people move through galleries, which exhibits attract the most attention, and how visitors interact with collections, museums can design exhibitions that are more intuitive, accessible, and enjoyable. These insights help institutions improve storytelling, enhance visitor engagement, and create experiences that appeal to audiences of all ages.

Technologies such as artificial intelligence, interactive displays, and digital guides are transforming museums into more immersive cultural spaces. Rather than replacing the traditional museum experience, these innovations complement it by making art, history, and heritage more accessible and interactive. The challenge lies in using technology responsibly, respecting visitors’ privacy while ensuring that digital tools enrich, rather than overshadow, the cultural and educational value of their collections.

More information:

https://arstechnica.com/culture/2026/07/with-help-from-data-art-museums-are-reframing-the-visitor-experience/

Snake Robots Aid Earthquake Rescue Efforts

Following the devastating earthquakes that struck northern Venezuela in June 2026, researchers from Carnegie Mellon University's Robotics Institute deployed advanced snake robots to assist international search-and-rescue teams in the heavily affected city of La Guaira. Designed to slither through narrow gaps in collapsed buildings, the robots are equipped with front-mounted cameras that allow rescuers to inspect areas inaccessible or too dangerous for humans. Working alongside the Venezuelan and Colombian Red Cross, Mexico's Topos rescue team, and volunteers from around the world, the CMU team aimed to improve the speed and safety of survivor searches.

Although the robots did not locate any survivors during the three-day mission, they played an important role by confirming that certain collapsed structures contained no trapped victims, allowing rescue teams to focus their efforts elsewhere and providing closure to waiting families. The deployment also highlighted both the promise and the practical challenges of using robotics in disaster response, including difficult terrain, equipment repairs, logistics, and communication barriers. The experience reinforced the potential of bio-inspired robotic systems as valuable tools that complement, rather than replace, human first responders during large-scale humanitarian emergencies.

More information:

https://www.cmu.edu/news/stories/archives/2026/july/snake-robots-support-earthquake-search-and-rescue-in-venezuela