Automation in Construction in brief – 15/02/2025
Automation and Digitization
The Paris AI Action Summit, held in early 2025, gathered global leaders to discuss artificial intelligence’s future. Key outcomes included commitments to advancing AI for public benefit, fostering innovation, and ensuring transparency. Topics like AI’s role in the workforce, trust in technology, and its governance were central. Participants pledged to develop regulations for ethical AI while enhancing global cooperation to address risks and unlock its potential. The summit emphasized AI as a tool for societal good and economic progress.
Source: https://www.elysee.fr/en/sommet-pour-l-action-sur-l-ia
Mistral AI’s chatbot, Le Chat, launched on February 6th, quickly becoming France’s top iOS download. Faster than ChatGPT and using open-source models, it avoids national-security concerns. President Macron championed it at Paris’s AI summit, where France announced €109bn in AI investment. While Mistral remains a minnow next to U.S. giants, Le Chat is making waves in Europe’s AI landscape.
Source: https://www.economist.com/europe/2025/02/13/is-frances-le-chat-in-fact-a-cat
Researchers from Stanford and the University of Washington have created an efficient new approach to language modeling, scaling up performance by using budget forcing to control test-time compute. Their model, s1-32B, outperforms OpenAI’s o1-preview by up to 27% on math-based tasks. By fine-tuning on just 1,000 curated questions, the model achieves strong reasoning results, demonstrating the potential of test-time scaling with minimal training data and computational resources.
Source: https://arxiv.org/pdf/2501.19393
According to The Economist AI was once seen as a great equalizer, but recent research suggests it will widen social divides. While AI boosts productivity for low-skilled workers in routine tasks, high performers in complex fields benefit far more. Elite researchers, investors, and entrepreneurs leverage AI to amplify their expertise, leaving others behind. As AI reshapes jobs, those with judgment and adaptability will thrive, while others risk obsolescence.
Source: https://www.economist.com/finance-and-economics/2025/02/13/how-ai-will-divide-the-best-from-the-rest
A recent study found biases in chatbot responses, focusing on public awareness and government actions while neglecting the role of business and investors. These biases could shape public discourse and influence policymaking, undermining more transformative approaches to environmental problems. AI-powered chatbots, when asked about environmental challenges, tend to propose modest, incremental solutions that reinforce the status quo, rather than radical shifts necessary to address the scale of issues like climate change.
Meta has faced allegations of using unauthorized books to train its AI models. Court filings reveal that CEO Mark Zuckerberg approved the use of pirated materials, including e-books from the LibGen database. The company is now embroiled in a lawsuit from authors claiming copyright infringement, raising concerns about the legality of training large language models on stolen intellectual property.
Researchers at Fraunhofer IFF are developing an automated process for disassembling electronic waste, with an emphasis on reducing both material and energy costs. The project, iDEAR, uses AI-driven 3D cameras and optical sensors to identify and assess components in devices such as PCs, with the goal of expanding to other electronics. The method promises to streamline the recycling process and facilitate the reuse of valuable parts, advancing the circular economy by reducing waste and engineering effort.
Source: https://www.vdi-nachrichten.com/technik/automation/elektrogeraete-automatisiert-demontieren/
Construction Robotics
Wen et al. propose the Non-Dominated Sorting Black-Winged Kite Algorithm (NSBKA) to enhance fleet management in urban drone logistics. By optimizing multi-UAV path planning, NSBKA balances energy efficiency, safety, and computational speed. Benchmark tests and simulations confirm its improvement over existing methods. However, as fleet size increases, computation times grow, posing scalability challenges. Future research must address this bottleneck to improve operational feasibility and large-scale coordination.
Source: https://www.mdpi.com/2504-446X/9/2/118
Recent developments in robotics focus on enhancing lifelike qualities by improving materials and artificial skin. Soft robotics, using flexible, stretchable materials, has led to better facial expressions and more fluid movements. Advanced sensors and AI algorithms are now capable of detecting touch and pressure, enabling more precise interactions. Additionally, electroactive polymers mimic human skin’s responsiveness, improving tactile sensations.
Source: https://scitechdaily.com/beyond-the-uncanny-valley-new-tech-makes-robots-more-lifelike/
Amazon is advancing its robotics initiatives to further automate its fulfillment centers. The new robots, such as Sequoia, Hercules, and Titan, integrate AI and computer vision to improve storage, navigation, and load handling. Amazon also plans to expand these systems across its network, aiming to enhance efficiency. However, while automation increases productivity, it raises concerns about potential job displacement, despite claims of growing headcount in tech and maintenance roles.
Source: https://www.machine.news/amazon-prepares-for-next-wave-of-its-robotics-revolution/
Zhang et al. address challenges in drone-based bridge inspections, particularly the issue of GNSS signal shadowing. They propose a path-planning method designed to mitigate obstacles and signal loss, improving drone navigation and inspection accuracy. By factoring in environmental variables, the study seeks to enhance both the safety and efficiency of bridge assessments, especially in areas with weak GPS coverage.
Source: https://www.mdpi.com/2504-446X/9/2/124
Haba et al. (2025) propose a quantum annealing-based approach to optimize routing and scheduling for urban air mobility (UAM) fleets. Using mathematical optimization, the study addresses the challenges of managing dense air traffic in cities. By formulating route planning as a maximum weighted independent set problem, their method leverages quantum annealing to improve airspace utilization, validated with a simulator for Singapore’s airspace. The research highlights the potential of quantum computing in enhancing UAM traffic management.
Space Construction
Scottish firm Orbex is advancing 3D-printed rocket technology for small satellite launches. Its 19m tall Prime rocket features fully additive-manufactured engines, eliminating welds and reducing material waste. Running on low-emission biopropane, it offers a greener alternative to traditional rockets. With launches set from Saxavord Spaceport, Orbex aims to establish the UK as a leader in sustainable and efficient spaceflight.
Source: https://orbex.space/news/uk-government-invests-in-uk-orbital-launch-capability-for-first-time
China has begun recruiting for a planetary defence force, responding to the increasing risk of asteroid impacts, including the 2024 YR4 asteroid, which has a growing likelihood of striking Earth in 2032. The focus is on monitoring and early-warning systems, with an eye toward developing asteroid-deflection technologies. This expertise will be crucial not only for Earth but also for lunar infrastructure, where vulnerability to smaller meteorites will be a significant challenge due to the lack of atmosphere.
Materialise, a leader in additive manufacturing, has opened an aerospace competence center in Delft, Netherlands, to bolster its role in the aerospace sector. The company, which has delivered over 500,000 3D-printed parts, aims to accelerate innovation and partnerships within the Aerospace Innovation Hub. With EASA approval and EN 9100 certification, Materialise supports aerospace manufacturers, airlines, and maintenance providers with sustainable, certified 3D-printed components.
