A fridge-sized satellite is set to change space tech by using onboard AI to monitor and predict its own power system health, cutting design and launch time to just 13 months. This leap isn’t just about speed but about smarter autonomy in orbit, which will reduce reliance on ground control and improve mission resilience. While some may doubt AI’s role in critical systems, this project shows cautious optimism is warranted when innovation meets rigorous engineering. It’s a clear sign that space hardware is evolving, and we should watch closely how autonomy reshapes the future of spacecraft design and operation.

Read the full article here

Google’s datacentres now demand immense power, so it’s investing broadly from fusion and fission to geothermal and solar to meet rising needs sustainably. While fusion remains experimental, backing startups like Commonwealth Fusion Systems shows confidence in long-term innovation rather than quick fixes. Expanding geothermal and securing large-scale solar supply complement this approach, which wisely balances cutting-edge tech with proven sources. It’s a pragmatic example of how industry leaders can tackle energy challenges without overpromising, and it hints at a future where diverse, reliable clean energy powers critical infrastructure. This kind of strategy deserves close attention from engineers and policymakers alike.

Read the full article here

China’s latest move with “Big Fund III”, pouring billions into lithography and EDA tool development, marks a sharp strategic turn that every hardware engineer should watch closely. I’m always fascinated by efforts to chip away at technological dependencies, and here, the sheer scale is as impressive as the ambition. While reinventing the wheel isn’t quick or cheap, China’s commitment to technological self-sufficiency tells us just how vital cutting-edge manufacturing tools have become. Whether you’re on the shop floor or at the drawing board, there’s plenty to learn from their whole-of-nation approach.

Read the full article here

Microsoft's decision to cut up to 9,000 jobs underscores how AI investment is reshaping even tech giants. It’s striking that while gaming divisions face cuts, the company pours $80 billion into AI infrastructure and talent, signaling where future growth lies. This shift, though tough on affected teams, highlights the relentless pace of innovation demanding resource realignment. For those of us who build the circuits and systems powering AI, it’s a reminder that adaptability is key, and that the interplay between hardware and AI software will define the next era in technology.

Read the full article here

When Xbox’s Matt Turnbull suggested using AI tools like ChatGPT to cope with job loss emotions, it sparked mixed reactions, understandable given the sensitivity. While AI can support tasks like career planning or resume writing, it can’t replace genuine human empathy during layoffs. Still, his approach highlights how AI is becoming intertwined with everyday challenges, beyond just engineering feats. As we navigate these shifts, balancing technology’s promise with real human impact will be crucial, especially in industries facing rapid change and tough decisions. This story reminds us that innovation isn’t just technical, it’s also deeply personal.

Read the full article here

The disruptive and costly process of digging up roads to fix pipes might soon be a thing of the past, thanks to the University of Sheffield’s innovative Pipebots. These small, agile robots navigate inside pipes detecting leaks with precision, and they even promise future in-pipe repairs without excavation. This tech isn’t just clever engineering; it addresses real-world challenges, aging infrastructure and escalating water waste, that many regions face globally. For those of us concerned with practical, effective solutions, Pipebots offer a promising glimpse of how robotics and AI can make infrastructure maintenance smarter and less invasive.

Read the full article here

China’s unveiling of a homegrown seven-function robotic arm for deep-sea oil and gas work is a notable engineering feat. At a maximum depth of 7,000 meters, this lightweight yet robust device performs complex tasks like valve operation and equipment installation under challenging conditions. What stands out isn’t just its capabilities, but also its 35% lighter weight and 40% cost savings compared to imported models. This move reflects a broader trend where domestic innovation reduces reliance on foreign tech, which is crucial in sensitive sectors like deep-sea energy. It’s a practical example of how engineering advances secure strategic industrial competitiveness.

Read the full article here

Researchers at University College London have made a significant stride with a gelatin-based robotic skin that senses touch, temperature, and damage simultaneously. Unlike traditional multi-sensor setups, this single-material skin offers robustness and cost-effectiveness, which is promising for prosthetics, humanoid robots, and safety applications. Testing involved everything from heat blasts to deep cuts, feeding millions of data points into AI models for nuanced touch recognition. While not yet matching human skin, this innovation brings us closer to robots that can truly perceive their environment, a crucial leap for hardware engineering where tactile feedback is essential for real-world interaction.

Read the full article here

A breakthrough from the University of Tsukuba could reshape solar tech by boosting tin-based perovskite solar cells, replacing toxic lead. Their research shows that incorporating indene-C60 diadduct (ICBA) into the electron transport layer improves efficiency by reducing charge recombination, raising open-circuit voltage. This addresses a key barrier for tin-based cells, which have lagged behind lead counterparts. Given the environmental and manufacturing advantages of tin perovskites (like lower toxicity and flexible, low-temp production), this advancement offers a promising path toward sustainable, cost-effective solar solutions for the future.

Read the full article here

Imagine a vision sensor that not only sees but adapts like the human eye, and even outperforms it. Researchers at Fuzhou University have developed such a breakthrough using lead sulfide quantum dots, enabling ultra-fast, energy-efficient light adaptation directly on the sensor. This neuromorphic design mimics retinal processes by trapping and releasing charge, filtering data before it reaches AI processors. For autonomous vehicles, drones, and edge AI devices, this means faster, smarter responses to rapidly changing light conditions without the heavy computational cost. It’s a compelling step toward machines that truly perceive their world, moving us closer to seamless human-machine interaction.

Read the full article here

A 160-core RISC-V cluster squeezed onto an M.2 board might sound surprising, yet [bitluni]’s recent project does just that using CH32V003 cores running at 48 MHz each. While raw clock speeds are modest, the massive parallelism recalls mid-80s supercomputing power, enough to demo a raymarcher, albeit far from RTX-level graphics. Communication over PCIe via a CH382 interface is currently the bottleneck, limiting data throughput despite ample compute capacity. This inventive approach shows how low-power RISC-V microcontrollers can form powerful coprocessors for specialized, compute-heavy tasks, hinting at creative new uses for embedded parallelism in hardware design.

Read the full article here