Small two-wheeled robots, equipped with high-tech sensors, will help to find survivors faster in the aftermath of disasters. © Tohoku University, 2023.

By Michael Allen

In the critical 72 hours after an earthquake or explosion, a race against the clock begins to find survivors. After that window, the chances of survival drop sharply.

When a powerful earthquake hit central Italy on 24 August 2016, killing 299 people, over 5 000 emergency workers were mobilised in search and rescue efforts that saved dozens from the rubble in the immediate aftermath.

The pressure to move fast can create risks for first responders, who often face unstable environments with little information about the dangers ahead. But this type of rescue work could soon become safer and more efficient thanks to a joint effort by EU and Japanese researchers.

Supporting first responders

Rescue organisations, research institutes and companies from both Europe and Japan worked together from 2019 to 2023 to develop a new generation of tools blending robotics, drone technology and chemical sensing to transform how emergency teams operate in disaster zones.

It is a prototype technology that did not exist before.
– Tiina Ristmäe, CURSOR

Their work was part of a four-year EU-funded international research initiative called CURSOR, which included partners from six EU countries, Norway and the UK. It also included Tohoku University, whose involvement was funded by the Japan Science and Technology Agency.

The researchers hope that the sophisticated rescue kit they have developed will help rescue workers locate trapped survivors faster, while also improving their own safety.

“In the field of search and rescue, we don’t have many technologies that support first responders, and the technologies that we do have, have a lot of limitations,” said Tiina Ristmäe, a research coordinator at the German Federal Agency for Technical Relief and vice president of the International Forum to Advance First Responder Innovation.

Meet the rescue bots

At the heart of the researcher’s work is a small robot called Soft Miniaturised Underground Robotic Finder (SMURF). The robot is designed to navigate through collapsed buildings and rubble piles to locate people who may be trapped underneath.

The idea is to allow rescue teams to do more of their work remotely, localising and finding humans from the most hazardous areas in the early stages of a rescue operation. The SMURF can be remotely controlled by operators who stay at a safe distance from the rubble.

“It is a prototype technology that did not exist before,” said Ristmäe. “We don’t send people, we send machines – robots – to do the often very dangerous job.”

The SMURF is compact and lightweight, with a two-wheel design that allows it to manoeuvre over debris and climb small obstacles.

“It moves and drops deep into the debris to find victims, with multiple robots covering the whole rubble pile,” said Professor Satoshi Tadokoro, a robotics expert at Tohoku University and one of the project’s lead scientists.

The development team tested many designs before settling on the final SMURF prototype.

“We investigated multiple options – multiple wheels or tracks, flying robots, jumping robots – but we concluded that this two-wheeled design is the most effective,” said Tadokoro.

Sniffing for survivors

The SMURF’s small “head” is packed with technology: video and thermal cameras, microphones and speakers for two-way communication, and a powerful chemical sensor known as the SNIFFER.

This sensor is capable of detecting substances that humans naturally emit, such as C02 and ammonia, and can even distinguish between living and deceased individuals.

Put to the test in real-world conditions, the SNIFFER has proved able to provide reliable information even when surrounded by competing stimuli, like smoke or rain.

According to the first responders who worked with the researchers, the information provided by the SNIFFER is highly valuable: it helps them to prioritise getting help to those who are still alive, said Ristmäe.

Drone delivery

To further improve the reach of the SMURF, the researchers also integrated drone support into the system. Customised drones are used to deliver the robots directly to the areas where they’re needed most – places that may be hard or dangerous to access on foot.

Ιt moves and drops deep into the debris to find victims, with multiple robots covering the whole rubble pile.
– Professor Satoshi Tadokoro, Tohoku University

“You can transport several robots at the same time and drop them in different locations,” said Ristmäe.

Alongside these delivery drones, the CURSOR team developed a fleet of aerial tools designed to survey and assess disaster zones. One of the drones, dubbed the “mothership,” acts as a flying communications hub, linking all the devices on the ground with the rescue team’s command centre.

Other drones carry ground-penetrating radar to detect victims buried beneath debris. Additional drones capture overlapping high-definition footage that can be stitched together into detailed 3D maps of the affected area, helping teams to visualise the layout and plan their operations more strategically.

Along with speeding up search operations, these steps should slash the time emergency workers spend in dangerous locations like collapsed buildings.

Testing in the field

The combined system has already undergone real-world testing, including large-scale field trials in Japan and across Europe.

One of the most comprehensive tests took place in November 2022 in Afidnes, Greece, where the full range of CURSOR technologies was used in a simulated disaster scenario.

Though not yet commercially available, the prototype rescue kit has sparked global interest.

“We’ve received hundreds of requests from people wanting to buy it,” said Ristmäe. “We have to explain it’s not deployable yet, but the demand is there.”

The CURSOR team hopes to secure more funding to further enhance the technology and eventually bring it to market, potentially transforming the future of disaster response.

Research in this article was funded by the EU’s Horizon Programme. The views of the interviewees don’t necessarily reflect those of the European Commission. If you liked this article, please consider sharing it on social media.

---

This article was originally published in Horizon, the EU Research and Innovation magazine.

Birds flock in order to forage and move more efficiently. Fish school to avoid predators. And bees swarm to reproduce. Recent advances in artificial intelligence have sought to mimic these natural behaviors as a way to potentially improve search-and-rescue operations or to identify areas of wildfire spread over vast areas—largely through coordinated drone or robotic movements. However, developing a means to control and utilize this type of AI—or "swarm intelligence"—has proved challenging.

The next era of robotics and AI won't be defined solely by technical breakthroughs but by how well these systems integrate into human environments. No matter how powerful the sensors or how technical the learning models are, without user trust, adoption will stall.

Artificial intelligence is consuming enormous amounts of energy, but researchers at the University of Florida have built a chip that could change everything by using light instead of electricity for a core AI function. By etching microscopic lenses directly onto silicon, they’ve enabled laser-powered computations that cut power use dramatically while maintaining near-perfect accuracy.

A team of scientists from China and the U.S. is pioneering the development of bubble-powered robots, which could one day replace needles for painless drug delivery into the body. Inspired by nature, the researchers developed a new technique that harnesses the energy released by a collapsing bubble in a liquid, a process known as cavitation.

In this conversation, Jonathan Tan explains how energy needs vary across robot categories, why current systems are holding back innovation, and what battery breakthroughs could mean for the future of robotics.

New AI Forges Video Stories in Minutes

AI startup Fable Studio is out with a new tool — dubbed ‘Showrunner,’ that long-term, promises to forge entire TV episodes and movies in minutes.

Simply enter a text prompt describing what you want, click enter, and in about 22 minutes, you’ll have your 22-minute TV episode.

Even better: You can use the tool to drop yourself into the action as a character.

Currently demoing with short clips in cartoon form, Fable Studio is promising film-like TV shows and movies in about two years, according to CEO Edward Saatchi.

In other news and analysis on AI:

*ChatGPT Rated Number One Consumer App: Venture capital firm Andreessen Horowitz has released its latest analysis of the AI market, pegging ChatGPT as the top AI app, based on unique monthly visits.

On its heels is Google’s Gemini chatbot, followed by DeepSeek – a chatbot made in China.

Also high-up on the list:

–Grok
–Character.ai
–Perplexity
–Claude

*Google’s Text-to-Talking-Heads Podcast Tool Gets an Upgrade: NotebookLM, a research and production tool that has wowed users by being able to transform an article or other text into an audio podcast featuring two people discussing the substance of that text has a new spring in its step.

Now, you can tweak the tool so that your text-to-podcast emerges in any of four formats:

–Brief, which offers-up a two-minute summary of the text
–Deep Dive, which offers an in-depth discussion of your
material
–Critique, which features talking heads critically discussing
your text
–Debate, which creates two hosts who take different points-
of-view on your article or similar

*Google’s New ‘Nano Banana’ Image Editor: Cool Use Cases: The fervor over Google’s new image editor continues to rage across the Web, as increasingly numbers of users are entranced by its power and surgical precision.

One of the new tool’s most impressive features: The ability to stay true to the identity of a human face – no matter how many times it remakes that image.

For a quick study, check-out these videos on YouTube, that show you scores of ways to use the new editor – officially known as Gemini 2.5 Flash Image:

–Google Gemini 2.5 Flash Image (Nano Banana) – 20 Creative Use Cases

–15 New Use Cases with Nano Banana

–The Ultimate Guide to Gemini 2.5- Flash (Nano Banana)

–New Gemini 2.5 Flash Image is Insane & Free

–Nano Banana Just Crushed Image Editing

*One Educator’s Take: Human-Generated Writing Still Essential in the Age of AI: College writing teacher Liz Stillwaggon Swan insists that without formal writing instruction, college students will be unleashed on the world sans the ability to think clearly and deeply.

Observes Swan: “I explain to my students that writing is a process of making the subconscious conscious—of bringing hazy, half-baked assumptions, biases, intuitions, ideas, anxieties, and hopes to the surface.

”Often, we don’t know what we believe until we start writing. We put our feelings and experiences into words and stories, even arguments, and through that arduous process, we begin to feel utterly human.”

*Another Educator’s Take: AI Has Rendered Traditional Writing Instruction Obsolete: It’s time to trash the teaching of writing at the college level as we know it, according to John Villasenor, a writing instructor at University of California Los Angeles.

Instead, today’s college students – who already know that AI will be handling most of the writing needed in years to come – should be taught how to get the most from AI when using it for writing.

Observes Villasenor: “It means helping students become proficient at using AI as a force multiplier to improve the depth, versatility, and speed of their writing.

“Today’s young people know that when it comes to writing, the technology landscape has undergone a tectonic shift, and they have already found their new footing. Those of us involved in teaching them need to do the same.”

*AI Agents and Marketing: A Primer: AI startup Smartcat.ai is offering a free eBook detailing how marketers can use multiple agents to automate much of their work.

Observes Nicole Di Nicola, VP of marketing, Smartcat.ai: “It’s like every marketer can now become a content creator. A product marketer can take a messaging doc and have AI turn it into a campaign plan with emails and sequences—fewer handoffs, less lag, more ownership. ”

One caveat: Given that AI agents are brand new technology that sometimes gets ahead of its skis, ‘pilot trial’ is the operative phrase here.

*Anthropic to Cough-Up $1.5 Billion to Book Authors: AI startup Anthropic – maker of the popular Claude chatbot – has agreed to pay $1.5 billion to authors and publishers as compensation for using their intellectual property to train its AI.

Observes writer Cade Metz: “The settlement is the largest payout in the history of U.S. copyright cases.

”Anthropic will pay $3,000 per work to 500,000 authors.”

*Google Promising AI Writing for More Android Phones: Google’s Gboard’s AI Writing Tools will be rolled-out to more phones in coming weeks, according to the tech goliath.

The tool enables users to proofread and rephrase text on their Android smartphones – without being forced to go to the cloud.

But so far, only devices featuring Gemini Nano v2 or higher are being promised the tools.

*AI BIG PICTURE: AI’s Next Killer App: Emotional Manipulation?: Geoffrey Hinton, the Godfather of AI, warns that the tech will soon be better at manipulating people than the most accomplished con man.

Observes writer Eric Hal Schwartz: Hinton “believes AI will be smarter than humans in ways that let them push our buttons, make us feel things, change our behavior, and do it better than even the most persuasive human being.

“The nightmare is an AI that understands us so well that it can change us, not by force, but by suggestion and influence.”

Share a Link:  Please consider sharing a link to https://RobotWritersAI.com from your blog, social media post, publication or emails. More links leading to RobotWritersAI.com helps everyone interested in AI-generated writing.

–Joe Dysart is editor of RobotWritersAI.com and a tech journalist with 20+ years experience. His work has appeared in 150+ publications, including The New York Times and the Financial Times of London.

The post Hollywood Killer appeared first on Robot Writers AI.

Artificial intelligence is reshaping law, ethics, and society at a speed that threatens fundamental human dignity. Dr. Maria Randazzo of Charles Darwin University warns that current regulation fails to protect rights such as privacy, autonomy, and anti-discrimination. The “black box problem” leaves people unable to trace or challenge AI decisions that may harm them.

A spinal cord injury in most vertebrates likely inhibits locomotion and induces paralysis—not so in eels. They not only possess the ability to move through water, and surprisingly, across land when intact, but can also continue to swim even if their spinal cord is severed.

Although regulating the safe use of fully autonomous vehicles across jurisdictions and international routes still has a fair bit of ironing out to do, AT providers and OEMs are making stark improvements to the technology inside our trucks.

Our novel Deep Loop Shaping method improves control of gravitational wave observatories, helping astronomers better understand the dynamics and formation of the universe.

Scientists at UCL, Google DeepMind and Intrinsic have developed a powerful new AI algorithm that enables large sets of robotic arms to work together faster and smarter in busy industrial settings, potentially saving manufacturers hundreds of hours of planning time and unlocking new levels of flexibility and efficiency.

Anyone at the show who visits our booth will be able to find something relevant to their own operations that can solve a key problem and offer improved performance, lower costs, and greater sustainability.

General Motors Co. worker Annie Ignaczak spent years walking in circles on concrete factory floors, assembling the same parts and counting down hundreds of pieces she and her coworkers needed to finish before lunch.

Image provided by the authors – generated using Gemini.

For many of us, artificial intelligence (AI) has become part of everyday life, and the rate at which we assign previously human roles to AI systems shows no signs of slowing down. AI systems are the crucial ingredients of many technologies — e.g., self-driving cars, smart urban planning, digital assistants — across a growing number of domains. At the core of many of these technologies are autonomous agents — systems designed to act on behalf of humans and make decisions without direct supervision. In order to act effectively in the real world, these agents must be capable of carrying out a wide range of tasks despite possibly unpredictable environmental conditions, which often requires some form of machine learning (ML) for achieving adaptive behaviour.

Reinforcement learning (RL) [6] stands out as a powerful ML technique for training agents to achieve optimal behaviour in stochastic environments. RL agents learn by interacting with their environment: for every action they take, they receive context-specific rewards or penalties. Over time, they learn behaviour that maximizes the expected rewards throughout their runtime.

Image provided by the authors – generated using Gemini.

RL agents can master a wide variety of complex tasks, from winning video games to controlling cyber-physical systems such as self-driving cars, often surpassing what expert humans are capable of. This optimal, efficient behaviour, however, if left entirely unconstrained, may turn out to be off-putting or even dangerous to the humans it impacts. This motivates the substantial research effort in safe RL, where specialized techniques are developed to ensure that RL agents meet specific safety requirements. These requirements are often expressed in formal languages like linear temporal logic (LTL), which extends classical (true/false) logic with temporal operators, allowing us to specify conditions like “something that must always hold”, or “something that must eventually occur”. By combining the adaptability of ML with the precision of logic, researchers have developed powerful methods for training agents to act both effectively and safely.

However, safety isn’t everything. Indeed, as RL-based agents are increasingly given roles that either replace or closely interact with humans, a new challenge arises: ensuring their behavior is also compliant with the social, legal and ethical norms that structure human society, which often go beyond simple constraints guaranteeing safety. For example, a self-driving car might perfectly follow safety constraints (e.g. avoiding collisions), yet still adopt behaviors that, while technically safe, violate social norms, appearing bizarre or rude on the road, which might cause other (human) drivers to react in unsafe ways.

Norms are typically expressed as obligations (“you must do it”), permissions (“you are permitted to do it”) and prohibitions (“you are forbidden from doing it”), which are not statements that can be true or false, like classical logic formulas. Instead, they are deontic concepts: they describe what is right, wrong, or permissible — ideal or acceptable behaviour, instead of what is actually the case. This nuance introduces several difficult dynamics to reasoning about norms, which many logics (such as LTL) struggle to handle. Even every-day normative systems like driving regulations can feature such complications; while some norms can be very simple (e.g., never exceed 50 kph within city limits), others can be more complex, as in:

1. Always maintain 10 meters between your vehicle and the vehicles in front of and behind you.
2. If there are less than 10 meters between you and the vehicle behind you, you should slow down to put more space between yourself and the vehicle in front of you.

(2) is an example of a contrary-to-duty obligation (CTD), an obligation you must follow specifically in a situation where another primary obligation (1) has already been violated to, e.g., compensate or reduce damage. Although studied extensively in the fields of normative reasoning and deontic logic, such norms can be problematic for many basic safe RL methods based on enforcing LTL constraints, as was discussed in [4].

However, there are approaches for safe RL that show more potential. One notable example is the Restraining Bolt technique, introduced by De Giacomo et al. [2]. Named after a device used in the Star Wars universe to curb the behavior of droids, this method influences an agent’s actions to align with specified rules while still allowing it to pursue its goals. That is, the restraining bolt modifies the behavior an RL agent learns so that it also respects a set of specifications. These specifications, expressed in a variant of LTL (LTLf [3]), are each paired with its own reward. The central idea is simple but powerful: along with the rewards the agent receives while exploring the environment, we add an additional reward whenever its actions satisfy the corresponding specification, nudging it to behave in ways that align with individual safety requirements. The assignment of specific rewards to individual specifications allows us to model more complicated dynamics like, e.g., CTD obligations, by assigning one reward for obeying the primary obligation, and a different reward for obeying the CTD obligation.

Still, issues with modeling norms persist; for example, many (if not most) norms are conditional. Consider the obligation stating “if pedestrians are present at a pedestrian crossing, THEN the nearby vehicles must stop”. If an agent were rewarded every time this rule was satisfied, it would also receive rewards in situations where the norm is not actually in force. This is because, in logic, an implication holds also when the antecedent (“pedestrians are present”) is false. As a result, the agent is rewarded whenever pedestrians are not around, and might learn to prolong its runtime in order to accumulate these rewards for effectively doing nothing, instead of efficiently pursuing its intended task (e.g., reaching a destination). In [5] we showed that there are scenarios where an agent will either ignore the norms, or learn this “procrastination” behavior, no matter which rewards we choose. As a result, we introduced Normative Restraining Bolts (NRBs), a step forward toward enforcing norms in RL agents. Unlike the original Restraining Bolt, which encouraged compliance by providing additional rewards, the normative version instead punishes norm violations. This design is inspired by the Andersonian view of deontic logic [1], which treats obligations as rules whose violation necessarily triggers a sanction. Thus, the framework no longer relies on reinforcing acceptable behavior, but instead enforces norms by guaranteeing that violations carry tangible penalties. While effective for managing intricate normative dynamics like conditional obligations, contrary-to-duties, and exceptions to norms, NRBs rely on trial-and-error reward tuning to implement norm adherence, and therefore can be unwieldy, especially when trying to resolve conflicts between norms. Moreover, they require retraining to accommodate norm updates, and do not lend themselves to guarantees that optimal policies minimize norm violations.

Our contribution

Building on NRBs, we introduce Ordered Normative Restraining Bolts (ONRBs), a framework for guiding reinforcement learning agents to comply with social, legal, and ethical norms while addressing the limitations of NRBs. In this approach, each norm is treated as an objective in a multi-objective reinforcement learning (MORL) problem. Reformulating the problem in this way allows us to:

• Prove that when norms do not conflict, an agent who learns optimal behaviour will minimize norm violations over time.
• Express relationships between norms in terms of a ranking system describing which norm should be prioritized when a conflict occurs.
• Use MORL techniques to algorithmically determine the necessary magnitude of the punishments we assign such that it is guarantied that so long as an agent learns optimal behaviour, norms will be violated as little as possible, prioritizing the norms with the highest rank.
• Accommodate changes in our normative systems by “deactivating” or “reactivating” specific norms.

We tested our framework in a grid-world environment inspired by strategy games, where an agent learns to collect resources and deliver them to designated areas. This setup allows us to demonstrate the framework’s ability to handle the complex normative scenarios we noted above, along with direct prioritization of conflicting norms and norm updates. For instance, the figure below

displays how the agent handles norm conflicts, when it is both obligated to (1) avoid the dangerous (pink) areas, and (2) reach the market (blue) area by a certain deadline, supposing that the second norm takes priority. We can see that it chooses to violate (1) once, because otherwise it will be stuck at the beginning of the map, unable to fulfill (2). Nevertheless, when given the possibility to violate (1) once more, it chooses the compliant path, even though the violating path would allow it to collect more resources, and therefore more rewards from the environment.

In summary, by combining RL with logic, we can build AI agents that do not just work, they work right.

This work won a distinguished paper award at IJCAI 2025. Read the paper in full: Combining MORL with restraining bolts to learn normative behaviour, Emery A. Neufeld, Agata Ciabattoni and Radu Florin Tulcan.

Acknowledgements

This research was funded by the Vienna Science and Technology Fund (WWTF) project ICT22-023 and the Austrian Science Fund (FWF) 10.55776/COE12 Cluster of Excellence Bilateral AI.

References

[1] Alan Ross Anderson. A reduction of deontic logic to alethic modal logic. Mind, 67(265):100–103, 1958.

[2] Giuseppe De Giacomo, Luca Iocchi, Marco Favorito, and Fabio Patrizi. Foundations for restraining bolts: Reinforcement learning with LTLf/LDLf restraining specifications. In Proceedings of the international conference on automated planning and scheduling, volume 29, pages 128–136, 2019.

[3] Giuseppe De Giacomo and Moshe Y Vardi. Linear temporal logic and linear dynamic logic on finite traces. In IJCAI, volume 13, pages 854–860, 2013.

[4] Emery Neufeld, Ezio Bartocci, and Agata Ciabattoni. On normative reinforcement learning via safe reinforcement learning. In PRIMA 2022, 2022.

[5] Emery A Neufeld, Agata Ciabattoni, and Radu Florin Tulcan. Norm compliance in reinforcement learning agents via restraining bolts. In Legal Knowledge and Information Systems JURIX 2024, pages 119–130. IOS Press, 2024.

[6] Richard S. Sutton and Andrew G. Barto. Reinforcement learning – an introduction. Adaptive computation and machine learning. MIT Press, 1998.