News

If startup success stories usually make you picture cutting-edge tech, you might be missing a big part of the picture. Sometimes, the weirdest ideas shine the brightest, making real money and delighting both founders and customers. From ordinary rocks turned into pets to renting live chickens, these unusual startups show how far creativity and a pinch of humor can go.  

If you think the business world is all suits and serious pitches, think again—welcome to the wonderfully weird side of entrepreneurship.

‍

1. Pet Rock: The low-maintenance pet

Owning a pet is a joy, but let’s be honest—it’s also a handful. Between shedding fur, endless feeding schedules, surprise messes and finding a sitter when you’re away, pet parenting is not exactly effortless.  

Back in 1975, an advertising executive named Gary Dahl found himself joking about this very problem over drinks with friends. His solution for the “perfect” pet: a rock. No feeding, no walking, no grooming and absolutely no accidents on the carpet.

What started as a joke quickly snowballed into a real business. Smooth stones were sourced from Rosarito Beach in Mexico, then packed in playful cardboard “pet carrier” boxes with little air holes and a bed of straw. To make the experience even more cheeky, every Pet Rock came with a care manual that instructed owners to give their new companion sunlight, affection and, of course, a name.

It was absurd and hilarious, but it worked. Selling at US$3.95 apiece in the ’70s, Pet Rocks became a cultural phenomenon. Today, you can still find them on Amazon, but they will now set you back around US$29.99 or more. Would you bring home a Pet Rock? People in the ’70s sure did.

‍

2. Rent The Chicken: Bringing fresh eggs home

Back in 2013, Phil and Jenn Tompkins a couple duo, launched the company "Rent The Chicken" with one straightforward goal: give people a chance to try raising backyard hens and enjoy fresh eggs without the long-term commitment.  

Through partnerships with local farmers across the U.S. and Canada, this backyard chicken rental startup brings egg-laying hens straight to people’s yards. It offers different rental packages, but a standard six-month rental costs around US$500. This usually includes two hens ready to lay within days, a portable coop, feed, food and water dishes and expert support for any chicken-related questions.

The chickens arrive in spring and stay until fall. When the season ends, families can choose to return the hens, extend the rental or even buy them for about US$40 each at the end of the contract.

Today, the company works with partners in 29 states, from Oregon to Texas, and in parts of Canadain p. For people outside those areas, an out-of-area purchase package that comes with three hens can be shipped anywhere in the 48 contiguous states in the U.S. for about US$1,550.

In a way, it’s a fun and hands-on path to food security — giving families the joy of collecting their own eggs and knowing exactly where their breakfast comes from.

‍

3. I Want to Draw a Cat For You: Personalized purr-fect artwork

By day, Gadlin worked as a full-time web developer for a television broadcasting company. Outside of work, he poured his energy into comedy and writing. That creative streak took him back to high school days, when he had drawn silly cats for a comic series called Silly Cats Comic.

With those doodles as his foundation and a bit of basic design know-how, Gadlin launched his website, “I Want to Draw a Cat For You” in 2011. The concept was as simple as it was funny: visitors would describe the cat of their dreams and Gadlin would personally hand-draw it, then send it their way.  

This quirky startup idea landed him on Shark Tank, where he secured an offer of US$25,000 from investor Mark Cuban for a 33% stake in the business. Not bad for stick-figure cats.

When the site first launched, customers could pay extra US$5 for colour. Shipping cost US$1 if they didn’t mind the drawing arriving in a folded envelope, or US$5 for a flat mailer. For delivery within 48 hours, there was a US$19.95 rush fee that many customers were happy to pay.  

These days, Gadlin leans more on digital delivery and limited runs of his cat drawings at US$50, rather than mailing every single piece of his art. What he once described as “mediocre cat drawings” has become proof that a simple, original idea can claw its way into the startup world.

‍

4. Rent-A-Friend: Friendly companionship for hire

Imagine arriving in a new city with no one to show you around. That is exactly the kind of situation where RentAFriend can help.  

Launched in 2009 by Scott Rosenbaum, the unusual business was inspired by Japan’s “rental family” services, where people can hire a friend, a date or even a parent for a short period. Rosenbaum saw an opportunity to adapt that concept for North America, but with a focus strictly on platonic friendship.

Here’s how it works: Anyone can sign up as a “friend” for free by creating a profile, listing their interests and setting an hourly rate. People who want to hire pay a membership fee, typically around US$24.95 a month, to connect with friends across the platform.  

With a rented friend, you can do pretty much anything platonic. Go sightseeing, hit a museum, catch a game, work out together or even bring them along to a party or family event. At its heart, RentAFriend connects people who need company with those happy to earn a little extra simply by being one.

‍

5. Entomo Farms: Farming crickets for protein

Back in 2014, in the small town of Norwood, Ontario, Canada, three brothers—Jarrod, Darren and Ryan Goldin, set out to do something that sounded downright bizarre at the time: farm crickets for people to eat.

The idea first struck Jarrod after he saw a cricket-based nutrition bar on television. Around the same time, the UN released a report on edible insects as a sustainable food source. Suddenly, the “weird” idea didn’t seem so weird after all.

At Entomo Farms, crickets are raised in cage-free “cricket condos”, where they live in warm, dark spaces that mimic their natural habitat. They’re fed and cared for until they reach about six or seven weeks old, then humanely harvested using a CO₂ method. From there, they’re rinsed, roasted and ground into a fine powder—no additives, just pure cricket protein.

The appeal goes beyond novelty. Crickets are packed with nutrients and need far less land, feed and water than beef, making them both healthy and eco-friendly.

While their approach may seem unconventional, what drives Entomo Farms is simple: making sustainable, responsible food accessible to everyone.

‍

Final thoughts

These startups prove that innovation doesn’t always wear a serious face. Sometimes, it turns up wrapped in humor, curiosity or even a touch of absurdity, yet still manages to spark real change. From crickets turned into protein to chickens rented out by the season, each weird startup idea shows that entrepreneurship thrives when people dare to think differently.

While some of these unusual business ideas burned bright then faded, others are still evolving in the background, shifting from fads to niche services or steady, quiet companies. What they share is a willingness to test an idea most people would dismiss at first glance.

That is the real takeaway for founders. Weird startup ideas will not always scale into unicorns, yet they can test new consumer habits, open up fresh markets and shape culture in surprising ways. If you are building something new, there is space for products that make people laugh, think or raise an eyebrow before they reach for their wallet.

Bindwell, a San Francisco–based ag-tech startup using AI to design new pesticide molecules, has raised US$6 million in seed funding, co-led by General Catalyst and A Capital, with participation from SV Angel and Y Combinator founder Paul Graham. The round will help the company expand its lab in San Carlos, hire more technical talent and advance its first pesticide candidates toward validation.  

Even as pesticide use has doubled over the last 30 years, farmers still lose up to 40% of global crops to pests and disease. The core issue is resistance: pests are adapting faster than the industry can update its tools. As a result, farmers often rely on larger amounts of the same outdated chemicals, even as they deliver diminishing returns.

Meanwhile, innovation in the agrochemical sector has slowed, leaving the industry struggling to keep up with rapidly evolving pests. This is the gap Bindwell is targeting. Instead of updating old chemicals, the company uses AI to find completely new compounds designed for today’s pests and farming conditions.  

This vision is made even more striking by the people leading it. Bindwell was founded by 18-year-old Tyler Rose and 19-year-old Navvye Anand, who met at the Wolfram Summer Research Program in 2023. Both had deep ties to agriculture — Rose in China and Anand in India — witnessing up close how pest outbreaks and chemical dependence burdened farmers.  

Filling the gap in today’s pesticide pipeline, Bindwell created an AI system that can design and evaluate new molecules long before they hit the lab. It starts with Foldwell, the company’s protein-structure model, which helps map the shapes of pest proteins so scientists know where a molecule should bind. Then comes PLAPT, which can scan through every known synthesized compound in just a few hours to see which ones might actually work. For biopesticides, they use APPT, a model tuned to spot protein-to-protein interactions and shown to outperform existing tools on industry benchmarks.

Bindwell isn’t selling AI tools. Instead, the company develops the molecules itself and licenses them to major agrochemical players. Owning the full discovery process lets the team bake in safety, selectivity and environmental considerations from day one. It also allows Bindwell to plug directly into the pipelines that produce commercial pesticides — just with a fundamentally different engine powering the science.

At present, the team is now testing its first AI-generated candidates in its San Carlos lab and is in early talks with established pesticide manufacturers about potential licensing deals. For Rose and Anand, the long-term vision is simple: create pest control that works without repeating the mistakes of the last half-century. As they put it, the goal is not to escalate chemical use but to design molecules that are more precise, less harmful and resilient against resistance from the start.

Hong Kong is entering a new phase of competition as mainland platforms accelerate their expansion into the city, turning it into a frontline testing ground for Chinese companies preparing to push into global markets. With retail, logistics and food-delivery businesses all reshaped in the past year, Hong Kong has become the closest international environment where mainland firms can experiment with pricing, supply chains and customer behaviour under a familiar regulatory and cultural framework.

The shift became especially clear this week. At HKTVmall’s Vision Day on November 11, 2025, CEO Ricky Wong warned that Hong Kong’s traditional retail model is facing its toughest moment yet. He said the biggest threat is not mainland competitors like Taobao, JD.com or Pinduoduo entering Hong Kong, but the city’s longstanding dependence on physical shopping. If local retailers do not evolve, he said, they risk becoming “very easy to die of thirst in the desert”. Wong even welcomed the rise of mainland e-commerce giants, arguing that the more players enter the city, the faster consumers will shift online — a transition HKTVmall relies on for growth.  

Yet his optimism is layered over a challenging reality. HKTVmall’s own numbers reflect pressure from competition and changing consumer habits. The company reported average daily GMV of HK$22.2 million during the latest shopping festival season — up 2.8% month-on-month but still down 4.3% compared year-on-year — showing that even established online platforms are struggling to maintain momentum as mainland entrants squeeze prices and widen product selection.

The city’s food-delivery market illustrates the shift even more sharply. Deliveroo, once the fastest-growing platform in Hong Kong and at one point holding more than half of the market, officially shut down in April this year after a long decline. Its trajectory mirrored the sector’s upheaval: the company surged during the pandemic but lost ground after restrictions eased, first overtaken by Foodpanda and then pressured heavily by Meituan-backed Keeta, which entered Hong Kong in 2023 and quickly seized about 30% of citywide orders.

Deliveroo’s exit and the handover of parts of its business to Foodpanda did little to stabilise the market. Keeta’s rapid expansion instead pushed Foodpanda onto the defensive, leaving two major players competing in a market shaped by mainland-style pricing and operations. Hong Kong’s delivery sector, once dominated by global firms, is increasingly defined by Chinese platforms optimizing speed and efficiency at a scale few competitors can match.

These changes are unfolding as Chinese companies shift their focus toward new global markets.  

With China reducing its reliance on the US and EU and exports steadily moving toward ASEAN, Hong Kong has become a strategic launchpad. The city’s proximity, language familiarity and regulatory structure make it the nearest international setting where Chinese firms can test overseas strategies before expanding into Southeast Asia, the Middle East or Latin America. The result is a competitive intensity that local companies have rarely experienced. Retailers face price pressure they can’t match, local platforms are losing ground to mainland giants and global players are struggling to stay in the game.

Consumers benefit from lower prices, faster delivery and wider choice — but for Hong Kong businesses, the landscape has turned unforgiving. Mainland companies are not treating Hong Kong as a final destination but as the first stop in a broader global push. That positioning is reshaping the city’s entire consumer economy. As more mainland firms look outward, Hong Kong’s role as a testing ground will only deepen and the first players to feel the impact will be those operating closest to the consumer.

In a world where digital borders are fading faster than ever, Tencent is betting on familiarity. With the launch of TenPay Global Checkout, the company wants to make paying across countries feel as seamless as paying at home.

The new service, unveiled today, allows Weixin Mini Program merchants outside mainland China to accept a variety of local payment methods. That includes digital wallets, real-time payment networks and credit and debit cards, all through a single integration. The launch starts in Singapore and Macao SAR, where merchants can now take payments via PayNow, BOCPAY(MO), and major cards. Japan, Australia and New Zealand are next, with more regions to follow soon.

This rollout builds on the growing reach of Weixin Mini Programs, known internationally through WeChat. These small apps are built right into the platform, letting users' shop, book services and make payments without downloading separate apps. Today, there are over one million monthly active users in key overseas markets, with Mini Programs available across 92 countries and regions.  

Yet, for many users abroad, paying within Mini Programs hasn’t always been simple. Foreign card restrictions, currency conversions and limited local options often made checkout a frustrating step. TenPay Global Checkout aims to change that.

“TenPay Global Checkout marks an important step in enhancing the local consumer experience. By enabling overseas Weixin Mini Program merchants to accept trusted and diversified local payment methods through one unified solution, users benefit from a more convenient and efficient payment experience.  This helps merchants improve payment conversion rates, expand their user base and scale their businesses to serve a broader range of customers”, said Wenhui Yang, CEO of TenPay Global (Singapore).

What makes this move interesting isn’t just its technical simplicity—it’s the cultural bridge it builds. For users in Singapore or Japan, paying with PayNow or a local card inside Weixin feels less like an international transaction and more like an everyday purchase.

For merchants, it’s an invitation into a market that values convenience and trust. Payment familiarity, after all, often decides whether a user completes a purchase or abandons it at checkout.

The company remains focused on creating secure, connected and user-friendly payment experiences that help merchants grow and allow consumers to pay with confidence, wherever they are.  

If successful, TenPay Global Checkout could quietly redefine how cross-border commerce feels—not like a transaction across regions, but a familiar tap, scan or click. In an increasingly global marketplace, that kind of familiarity might just be the next frontier in digital trust.

It’s no surprise that artificial intelligence, while remarkably capable, can also go astray—spinning convincing but entirely fabricated narratives. From politics to academia, AI’s “hallucinations” have repeatedly shown how powerful technology can go off-script when left unchecked.

Take Grok-2, for instance. In July 2024, the chatbot misled users about ballot deadlines in several U.S. states, just days after President Joe Biden dropped his re-election bid against former President Donald Trump. A year earlier, a U.S. lawyer found himself in court for relying on ChatGPT to draft a legal brief—only to discover that the AI tool had invented entire cases, citations and judicial opinions. And now, the academic world has its own cautionary tale.

Recently, a journal paper from the Department of Social Work and Social Administration at the University of Hong Kong was found to contain fabricated citations—sources apparently created by AI. The paper, titled “Forty Years of Fertility Transition in Hong Kong,” analyzed the decline in Hong Kong’s fertility rate over the past four decades. Authored by doctoral student Yiming Bai, along with Yip Siu-fai, Vice Dean of the Faculty of Social Sciences and other university officials, the study identified falling marriage rates as a key driver behind the city’s shrinking birth rate. The authors recommended structural reforms to make Hong Kong’s social and work environment more family-friendly.

But the credibility of the paper came into question when inconsistencies surfaced among its references. Out of 61 cited works, some included DOI (Digital Object Identifier) links that led to dead ends, displaying “DOI Not Found.” Others claimed to originate from academic journals, yet searches yielded no such publications.

Speaking to HK01, Yip acknowledged that his student had used AI tools to organize the citations but failed to verify the accuracy of the generated references. “As the corresponding author, I bear responsibility”, Yip said, apologizing for the damage caused to the University of Hong Kong and the journal’s reputation. He clarified that the paper itself had undergone two rounds of verification and that its content was not fabricated—only the citations had been mishandled.

Yip has since contacted the journal’s editor, who accepted his explanation and agreed to re-upload a corrected version in the coming days. A formal notice addressing the issue will also be released. Yip said he would personally review each citation “piece by piece” to ensure no errors remain.

As for the student involved, Yip described her as a diligent and high-performing researcher who made an honest mistake in her first attempt at using AI for academic assistance. Rather than penalize her, Yip chose a more constructive approach, urging her to take a course on how to use AI tools responsibly in academic research.

Ultimately, in an age where generative AI can produce everything from essays to legal arguments, there are two lessons to take away from this episode. First, AI is a powerful assistant, but only that. The final judgment must always rest with us. No matter how seamless the output seems, cross-checking and verifying information remain essential. Second, as AI becomes integral to academic and professional life, institutions must equip students and employees with the skills to use it responsibly. Training and mentorship are no longer optional; they’re the foundation for using AI to enhance, not undermine, human work.

Because in this age of intelligent machines, staying relevant isn’t about replacing human judgment with AI, it’s about learning how to work alongside it.

Neuron7.ai, a company that builds AI systems to help service teams resolve technical issues faster, has launched Neuro. It is a new kind of AI agent built for environments where accuracy matters more than speed. From manufacturing floors to hospital equipment rooms, Neuro is designed for situations where a wrong answer can halt operations.

What sets Neuro apart is its focus on reliability. Instead of relying solely on large language models that often produce confident but inaccurate responses, Neuro combines deterministic AI — which draws on verified, trusted data — with autonomous reasoning for more complex cases. This hybrid design helps the system provide context-aware resolutions without inventing answers or “hallucinating”, a common issue that has made many enterprises cautious about adopting agentic AI.

“Enterprise adoption of agentic AI has stalled despite massive vendor investment. Gartner predicts 40% of projects will be canceled by 2027 due to reliability concerns”, said Niken Patel, CEO and Co-Founder of Neuron7. “The root cause is hallucinations. In service operations, outcomes are binary. An issue is either resolved or it is not. Probabilistic AI that is right only 70% of the time fails 30% of your customers and that failure rate is unacceptable for mission-critical service”.

That concern shaped how Neuro was built. “We use deterministic guided fixes for known issues. No guessing, no hallucinations — and reserve autonomous AI reasoning for complex scenarios. What sets Neuro apart is knowing which mode to use. While competitors race to make agents more autonomous, we're focused on making service resolution more accurate and trusted”, Patel explained.

At the heart of Neuro is the Smart Resolution Hub, Neuron7’s central intelligence layer that consolidates service data, knowledge bases and troubleshooting workflows into one conversational experience. This means a technician can describe a problem — say, a diagnostic error in an MRI scanner — and Neuro can instantly generate a verified, step-by-step solution. If the problem hasn’t been encountered before, it can autonomously scan through thousands of internal and external data points to identify the most likely fix, all while maintaining traceability and compliance.

Neuro’s architecture also makes it practical for real-world use. It integrates seamlessly with enterprise systems such as Salesforce, Microsoft, ServiceNow and SAP, allowing companies to embed it within their existing support operations. Early users of Neuron7’s platform have reported measurable improvements — faster resolutions, higher customer satisfaction and reduced downtime — thanks to guided intelligence that scales expert-level problem solving across teams.

The timing of Neuro’s debut feels deliberate. As organizations look to move past the hype of generative AI, trust and accountability have become the new benchmarks. AI systems that can explain their reasoning and stay within verifiable boundaries are emerging as the next phase of enterprise adoption.

“The market has figured out how to build autonomous agents”, Patel said. “The unsolved problem is building accurate agents for contexts where errors have consequences. Neuro fills that gap”.

Neuron7 is building a system that knows its limits — one that reasons carefully, acts responsibly and earns trust where it matters most. In a space dominated by speculation, that discipline may well redefine what “intelligent” really means in enterprise AI.

Encyclopaedias have always been mirrors of their time — from heavy leather-bound volumes in the 19th century to Wikipedia’s community-edited pages online. But as the world’s information multiplies faster than humans can catalogue it, even open platforms struggle to keep pace. Enter Botipedia, a new project from INSEAD, The Business School for the World, that reimagines how knowledge can be created, verified and shared using artificial intelligence.

At its core, Botipedia is powered by proprietary AI that automates the process of writing encyclopaedia entries. Instead of relying on volunteers or editors, it uses a system called Dynamic Multi-method Generation (DMG) — a method that combines hundreds of algorithms and curated datasets to produce high-quality, verifiable content. This AI doesn’t just summarise what already exists; it synthesises information from archives, satellite feeds and data libraries to generate original text grounded in facts.

What makes this innovation significant is the gap it fills in global access to knowledge. While Wikipedia hosts roughly 64 million English-language entries, languages like Swahili have fewer than 40,000 articles — leaving most of the world’s population outside the circle of easily available online information. Botipedia aims to close that gap by generating over 400 billion entries across 100 languages, ensuring that no subject, event or region is overlooked.

"We are creating Botipedia to provide everyone with equal access to information, with no language left behind", says Phil Parker, INSEAD Chaired Professor of Management Science, creator of Botipedia and holder of one of the pioneering patents in the field of generative AI. "We focus on content grounded in data and sources with full provenance, allowing the user to see as many perspectives as possible, as opposed to one potentially biased source".

Unlike many generative AI tools that depend on large language models (LLMs), Botipedia adapts its methods based on the type of content. For instance, weather data is generated using geo-spatial techniques to cover every possible coordinate on Earth. This targeted, multi-method approach helps boost both the accuracy and reliability of what it produces — key challenges in today’s AI-driven content landscape.

Additionally, the innovation is also energy-efficient. Its DMG system operates at a fraction of the processing power required by GPU-heavy models like ChatGPT, making it a sustainable alternative for large-scale content generation.

By combining AI precision, linguistic inclusivity and academic credibility, Botipedia positions itself as more than a digital library — it’s a step toward universal, unbiased access to verified knowledge.

"Botipedia is one of many initiatives of the Human and Machine Intelligence Institute (HUMII) that we are establishing at INSEAD", says Lily Fang, Dean of Research and Innovation at INSEAD. "It is a practical application that builds on INSEAD-linked IP to help people make better decisions with knowledge powered by technology. We want technologies that enhance the quality and meaning of our work and life, to retain human agency and value in the age of intelligence".

By harnessing AI to bridge gaps of language, geography and credibility, Botipedia points to a future where access to knowledge is no longer a privilege, but a shared global resource.

Artificial intelligence is beginning to change how scientists understand cancer at the cellular level. In a new collaboration, Bio-Techne Corporation, a global life sciences tools provider, and Nucleai, an AI company specializing in spatial biology for precision medicine, have unveiled data from the SECOMBIT clinical trial that could reshape how doctors predict cancer treatment outcomes. The results, presented at the Society for Immunotherapy of Cancer (SITC) 2025 Annual Meeting, highlight how AI-powered analysis of tumor environments can reveal which patients are more likely to benefit from specific therapies.

Led in collaboration with Professor Paolo Ascierto of the University of Napoli Federico II and Istituto Nazionale Tumori IRCCS Fondazione Pascale, the study explores how spatial biology — the science of mapping where and how cells interact within tissue — can uncover subtle immune behaviors linked to survival in melanoma patients.

Using Bio-Techne’s COMET platform and a 28-plex multiplex immunofluorescence panel, researchers analyzed 42 pre-treatment biopsies from patients with metastatic melanoma, an advanced stage of skin cancer. Nucleai’s multimodal AI platform integrated these imaging results with pathology and clinical data to trace patterns of immune cell interactions inside tumors.

The findings revealed that therapy sequencing significantly influences immune activity and patient outcomes. Patients who received targeted therapy followed by immunotherapy showed stronger immune activation, marked by higher levels of PD-L1+ CD8 T-cells and ICOS+ CD4 T-cells. Those who began with immunotherapy benefited most when PD-1+ CD8 T-cells engaged closely with PD-L1+ CD4 T-cells along the tumor’s invasive edge. Meanwhile, in patients alternating between targeted and immune treatments, beneficial antigen-presenting cell (APC) and T-cell interactions appeared near tumor margins, whereas macrophage activity in the outer tumor environment pointed to poorer prognosis.

“This study exemplifies how our innovative spatial imaging and analysis workflow can be applied broadly to clinical research to ultimately transform clinical decision-making in immuno-oncology”, said Matt McManus, President of the Diagnostics and Spatial Biology Segment at Bio-Techne.

The collaboration between the two companies underscores how AI and high-plex imaging together can help decode complex biological systems. As Avi Veidman, CEO of Nucleai, explained, “Our multimodal spatial operating system enables integration of high-plex imaging, data and clinical information to identify predictive biomarkers in clinical settings. This collaboration shows how precision medicine products can become more accurate, explainable and differentiated when powered by high-plex spatial proteomics – not limited by low-plex or H&E data alone”.

Dr. Ascierto described the SECOMBIT trial as “a milestone in demonstrating the possible predictive power of spatial biomarkers in patients enrolled in a clinical study”.

The study’s broader message is clear: understanding where immune cells are and how they interact inside a tumor could become just as important as knowing what they are. As AI continues to map these microscopic landscapes, oncology may move closer to genuinely personalized treatment — one patient, and one immune network, at a time.

Analog Devices (ADI), a global semiconductor company, launched CodeFusion Studio™ 2.0 on November 3, 2025. The new version of its open-source development platform is designed to make it easier and faster for developers to build AI-powered embedded systems that run on ADI’s processors and microcontrollers.

“The next era of embedded intelligence requires removing friction from AI development”, said Rob Oshana, Senior Vice President of the Software and Digital Platforms group at ADI. “CodeFusion Studio 2.0 transforms the developer experience by unifying fragmented AI workflows into a seamless process, empowering developers to leverage the full potential of ADI's cutting-edge products with ease so they can focus on innovating and accelerating time to market”.

The upgraded platform introduces new tools for hardware abstraction, AI integration and automation. These help developers move more easily from early design to deployment.

CodeFusion Studio 2.0 enables complete AI workflows, allowing teams to use their own models and deploy them on everything from low-power edge devices to advanced digital signal processors (DSPs).

Built on Microsoft Visual Studio Code, the new CodeFusion Studio offers built-in checks for model compatibility, along with performance testing and optimization tools that help reduce development time. Building on these capabilities, a new modular framework based on Zephyr OS lets developers test and monitor how AI and machine learning models perform in real time. This gives clearer insight into how each part of a model behaves during operation and helps fine-tune performance across different hardware setups.

Additionally, the CodeFusion Studio System Planner has also been redesigned to handle more device types and complex, multi-core applications. With new built-in diagnostic and debugging features — like integrated memory analysis and visual error tracking — developers can now troubleshoot problems faster and keep their systems running more efficiently.

This launch marks a deeper pivot for ADI. Long known for high-precision analog chips and converters, the company is expanding its edge-AI and software capabilities to enable what it calls Physical Intelligence — systems that can perceive, reason, and act locally.  

“Companies that deliver physically aware AI solutions are poised to transform industries and create new, industry-leading opportunities. That's why we're creating an ecosystem that enables developers to optimize, deploy and evaluate AI models seamlessly on ADI hardware, even without physical access to a board”, said Paul Golding, Vice President of Edge AI and Robotics at ADI. “CodeFusion Studio 2.0 is just one step we're taking to deliver Physical Intelligence to our customers, ultimately enabling them to create systems that perceive, reason and act locally, all within the constraints of real-world physics”.

Shanghai-based robotics firm AgiBot has taken a major step toward bringing artificial intelligence into real manufacturing. The company announced that its Real-World Reinforcement Learning (RW-RL) system has been successfully deployed on a pilot production line run in partnership with Longcheer Technology.  It marks one of the first real applications of reinforcement learning in industrial robotics.

The project represents a key shift in factory automation. For years, precision manufacturing has relied on rigid setups: robots that need custom fixtures, intricate programming and long calibration cycles. Even newer systems combining vision and force control often struggle with slow deployment and complex maintenance. AgiBot’s system aims to change that by letting robots learn and adapt on the job, reducing the need for extensive tuning or manual reconfiguration.

The RW-RL setup allows a robot to pick up new tasks within minutes rather than weeks. Once trained, the system can automatically adjust to variations, such as changes in part placement or size tolerance, maintaining steady performance throughout long operations. When production lines switch models or products, only minor hardware tweaks are needed. This flexibility could significantly cut downtime and setup costs in industries where rapid product turnover is common.

The system’s main strengths lie in faster deployment, high adaptability and easier reconfiguration. In practice, robots can be retrained quickly for new tasks without needing new fixtures or tools — a long-standing obstacle in consumer electronics production. The platform also works reliably across different factory layouts, showing potential for broader use in complex or varied manufacturing environments.

Beyond its technical claims, the milestone demonstrates a deeper convergence between algorithmic intelligence and mechanical motion.Instead of being tested only in the lab, AgiBot’s system was tried in real factory settings, showing it can perform reliably outside research conditions.

This progress builds on years of reinforcement learning research, which has gradually pushed AI toward greater stability and real-world usability. AgiBot’s Chief Scientist Dr. Jianlan Luo and his team have been at the forefront of that effort, refining algorithms capable of reliable performance on physical machines. Their work now underpins a production-ready platform that blends adaptive learning with precision motion control — turning what was once a research goal into a working industrial solution.

Looking forward, the two companies plan to extend the approach to other manufacturing areas, including consumer electronics and automotive components. They also aim to develop modular robot systems that can integrate smoothly with existing production setups.