« All Events

Launch event for the White Paper on Robotics in Construction

robotics • AI • Data construction • smart facade technologies • mobile platform • end-effector • Green Deal • Artificial Intelligence • Data • recycling • automation 

April 9 All day

El 9 de abril de 2026, la Plataforma Tecnológica Española de la Construcción (PTEC)  y la Plataforma Tecnológica Español de Robótica (HISPAROB) organizaron conjuntamente la presentación del Libro Blanco de Robótica en la Construcción, una jornada que reunió a expertos del ámbito tecnológico, empresarial e institucional para analizar el impacto de la robótica y la automatización en el futuro del sector.

El evento se celebró en el Sala A del Ministerio de Ciencia, Innovación y Universidades y contó con la participación de representantes de la administración pública, empresas líderes, centros tecnológicos y universidades, tanto del ámbito de la construcción como de la robótica

Discover more

DATE & TIME

April 9

WEBSITE

CHECK HERE FOR MORE INFO
Close

REGULUS

Construction robots lack perception achieving sub-cm 3-D understanding; current tools stay at cm accuracy, limiting performance. REGULUS— named after the brightest star in Leo constellation —is a hardware-agnostic perception module equipping any ground robot with “eyes and cognition,” set to redefine how machines perceive construction sites. With standardised power and data interfaces, REGULUS fits Space Mobilityʼs robots—LEO quadruped (launching Q1 2026) and wheeled robots already validated in PV parks and mines—and AMALTEA platforms. Achieving real-time localization, semantic segmentation and dense 3-D mapping with ≤ 5 mm accuracy in unstructured sites, REGULUS turns costly survey-grade technologies into an open, scalable European standard for sustainable automation. It fuses structured-light RGB-D, 360° and high-precision LiDAR (e.g., Ouster OS0-128 Rev.7 ≈ €17 k), inertial sensors and total-station anchors into a unified RGB point cloud referenced to the building frame. The semantic catalogue recognises supports, panels, AprilTags, expandable within the AMALTEA ecosystem. LiDAR provides wide awareness, steerable visual heads capture close detail, micro-stops (< 1 s) refine alignment, and a web HMI communicates actionable information. TRL path & heritage. Founders achieved 1 mm Airbus-fuselage alignment [1] and 3-D vineyard reconstruction with custom odometry and AprilTag loop closure [2][3], attaining sub-cm accuracy – core to REGULUS fusion. Space Mobility develops PERCEIVE, a funded AI-inspection system already validated in PV parks for anomaly detection, extended to mine inspection in darkness (contract pending) and supporting ESA-BIC Greece LEO operations in lunar-analogue terrains. 

Close

LCI-Pilot

LCI-Pilot addresses Challenge 8 – Life Cycle and Cost Assessment under the Software Solutions track. It tackles the major bottleneck of slow, inconsistent LCI/LCC generation for façade systems, where current manual processing of EPDs and technical files delays sustainability analysis by weeks and limits reuse scenarios. The project delivers a human-centred AI LCI assistant that couples Large Language Models with retrieval-augmented generation and structured parsers to automatically extract, normalise, and complete façade-specific life-cycle data. Outputs comply with ISO 14040/14044 and EN 15804, include provenance and confidence metadata, and are exported as machine-readable records interoperable with DT3 (Extended Digital Twin) and DT10 (Material Sorting and Recycling) for lifecycle tracking and circularity evaluation. Links to DT2 and DT9 enable dynamic updates from design to operation. A guided web interface keeps experts “in the loop,” ensuring explainability, auditability, and user control in line with AMALTEAʼs SSH and Responsible AI principles.  

By automating verifiable, auditable LCI/LCC generation with explainable AI and human oversight, LCI-Pilot strengthens AMALTEAʼs Digital Twin ecosystem, enabling data-driven, transparent sustainability decisions and contributing directly to EU Green Deal and Renovation Wave objectives. 

Close

ANCHOR

ANCHOR’s goal is to address the AMALTEA Challenge 7. The robotic installation of curtain-wall facades in dynamic construction sites still faces three main high-level challenges:  

  • Centimeter-range drift in sites missing absolute positioning systems;
  • frequent occlusions and layout changes that can compromise and invalidate pre-planned paths; and
  • limited human–machine interfaces (HMI) that hinder the operation and increase collision risk.  

To address these challenges, the innovative approach of ANCHOR will be steered by the following three objectives to advance from TRL 4 to TRL 6: O1 Develop Multi-sensor localization & exact re-localization tools: Unify GNSS-RTK (outdoors) with LiDAR/vision/IMU (indoors) in a graph-SLAM augmented with anchor factors. Perform registration of the point cloud and visual landmark with the BIM model and enable mm-level re-localization via visual fiducials/UWB/total-station observations for a precise approach (KPI1: Localization error<5mm). O2 Develop Dynamic scene understanding, predictive safety & robust navigation tools: Explore multi-layer navigation maps (occupancy, traversability) where BIM semantics such as walls, openings, and keep-out zones are projected into navigation layers to guide path planning and final pose approach for facade modules and apply predictive collision prevention with recovery behaviors. Incorporate signal-quality detection to maintain safe motion in clutter and occluded areas (KPI2: Autonomous A→B mission success rate ≥ 90% in mapped work zones with zero contact events). O3 Implement Operator-first HMI and workflow control/monitoring tools: Provide a tablet UI for managing autonomous and assisted teleoperation modes with live pose, map layers and BIM overlays, predicted paths, and programmable 2D/3D safety zones, enabling data logging for validation and pilot reporting. 

Close

AIM-Fusion

AIM-Fusion represents a paradigm-shifting approach to Challenge #6, developing an AI-Agent Modular Sensor-Fusion System that unifies LiDAR, RGB-D and total-station data into a real-time, colourised 3D point cloud within a ROS2-native, open, and standard-compliant architecture. It aims to establish a new European benchmark for trustworthy, high-precision robotic installation—delivering sub-centimetre accuracy, five-minute self-calibration, and full interoperability with AMALTEA Digital Tools. The project begins at TRL 5 (laboratory-validated fusion) and advances to TRL 7 (operational prototype) through three stages:   

  • architecture and calibration-workflow design,  
  • AI-enhanced fusion engine integration within ROS2, and  
  • field validation under real installation conditions at euroDAOʼs Montpellier test site, supported by research facilities available through its CIFRE collaboration with LIRMM (Université de Montpellier) and industrial partnerships providing operational data and safety-compliance frameworks. 

Beyond classical automation, AIM-Fusion leverages euroDAOʼs UltrathinkTM AI-Agent architecture, where each Agent acts as a virtual robotics engineer executing calibration, dataset curation and continuous ROS2 testing in the cloud. This distributed Agent workforce provides unlimited scalability and 24/7 selfoptimisation, ensuring a continuously improving system without human fatigue or downtime. In doing so, AIM-Fusion redefines human–AI collaboration in robotic construction, advancing a human-centred, transparent and sustainable AI paradigm fully aligned with the EU AI Act and the AMALTEA mission to build open, safe and interoperable technologies for Europeʼs digital-construction future. 

Close

CIRMA

CIRMA addresses Challenge 5: Adaptive manufacturing and quality control, targeting the technical obstacle of producing high-quality facade modules with arbitrary geometries without relying on predefined CAD files. Current approaches require rigid positioning and manual adjustment, slowing production and increasing errors. Our innovative approach leverages an AI-driven, environment-agnostic robotic platform capable of welding, sealing and in-line quality control. The robot reconstructs its surroundings in real time using 3D point cloud scanning with LiDAR, stereo/depth cameras, and generates adaptive toolpaths based on AI vision and historical production data. It dynamically adapts to obstacles using autonomous navigation with MoveIt in ROS2, and ensures operator safety via AI-based human detection triggering emergency stops. CIRMA integrates robotics, sensor fusion, AI, computer vision and adaptive control, enabling autonomous fabrication, quality control and safety in one solution. The project starts at TRL 5, with an adaptive welding solution employing real-time 3D reconstruction, obstacle-free trajectories without CAD, a welding end-effector and an intuitive GUI successfully demonstrated in controlled environments for shipbuilding welding tasks. Through deployment in a real operational façade manufacturing line (Pilot 2), CIRMA will achieve TRL7 by integrating its robotic and AI components with the AMALTEA Digital Tools and validating the solutionʼs performance, precision and scalability under real industrial conditions. The system will evolve from operator-assisted adaptive welding to AI-enabled path generation with human safety and real-time control, ensuring consistent and high-quality output for diverse geometries. 

Close

RoboInspect

The RoboInspect solution directly addresses Challenge 4 – Advanced Inspection and Control Systems for Quality Assessment in Robotic Manufacturing, by introducing a dual-sensor triangulation laser vision system coupled with AI-based control algorithms for façade module production. In current manufacturing lines, quality control for welding and structural silicone bonding is mainly manual, slow, and inconsistent. RoboInspect replaces this with a real-time, precision-grade inspection and feedback loop integrated into AMALTEAʼs Digital Solution for Manufacturing (DT5–DT7). The system employs two triangulation laser sensors:  

  • Pre-process sensor (ahead of the torch) – performs incremental 3D scanning of the joint to guide adaptive trajectory planning and process parameter optimization;  
  • Post-process sensor (after the torch) – inspects the completed weld or silicone bead, generating dense 3D profiles (±0.05 mm accuracy) for immediate defect detection and classification.  

Incremental point-cloud and mesh-generation algorithms allow continuous geometry reconstruction synchronized with the robot motion, enabling instantaneous quality evaluation without interrupting the process. AI/ML models detect discontinuities, pores, or bonding gaps and trigger automatic parameter corrections through ROS 2 / OPC UA interfaces connected to DT5–DT6. The collected inspection data are transferred to DT7 – AI-enhanced quality control of the façade manufacturing and optionally to DT3 – Extended Digital Twin for lifecycle traceability and predictive maintenance. This approach fulfils the Challenge 4 objectives by ensuring zero-defect manufacturing, shorter inspection cycles, and higher repeatability, supporting the AMALTEA goal of energy-efficient, low-waste façade production, waste reduction (–70%) and CO2 reduction. 

Close

PARAFORGE

PARAFORGE introduces an interoperable parametric design framework that connects real-time data from design, manufacturing, and recycling phases into a unified, human-centered workflow. Built on SEAMLEXITYʼs close collaboration with contractors, fabricators, and material suppliers, PARAFORGE embeds real-world process and material intelligence directly into the design environment to ensure both technical and environmental accuracy. Through seamless interoperability with the AMALTEA consortiumʼs infrastructure, the system enables interactive control of design parameters and live performance feedback, ensuring that the user remains at the core of every design decision. This approach acknowledges that automation alone is not the solution. True innovation lies in human–machine collaboration, where AI augments the designerʼs ability to make informed and creative choices. The proposed solution interfaces with DT1 (AI-enhanced parametric design system) and DT2 (simulation framework) to support large-scale dataset generation for AI training, while connecting downstream with DT4 (AI for manufacturing optimization), DT7 (AI quality control), and DT10 (material sorting and recycling) for lifecycle consistency. 

Close

FACET

This solution addresses challenge number 3: developing a web-based GUI to support better design decisions. The core goal is to deliver a human-centered, platform-independent interface that enables users to explore outputs from Multi-Objective Optimization (MOO) simulations. Our system puts designers, engineers, and decision-makers at the forefront. It allows them to interact with data-rich models and explore exported design iterations, tightly coupling 3D geometry with numerical and textual data from MOO (Multi-objective optimization) analytics. Users can engage with the data both manually and with assistance from our built-in AI RAG (Retrieval-Augmented Generation) assistant. Key features include:  

  • Interactive visualization and filtering of MOO results using multiple plots and charts (e.g., dual axis trade-offs, parallel coordinates).
  • Natural language queries (e.g., “show me solutions with R-value above 2.3 m2K/W and carbon footprint below 300 kgCO2e/m2) that the RAG system translates into executable filters, providing data provenance and contextual suggestions.
  • Ability to mark favorite solutions and track the design space.  
  • In-depth inspection of designs in 3D, including all layers and embedded metadata. 
Close

iFORGE

iFORGE translates novel existing I3D Robotics (i3D) technology to support Challenge#4 & Pilot testing by augmenting human skill with accurate, safe, collaborative robotics operating in real-time (RT). I3D designs, manufactures & integrates 3D machine vision systems utilising AI & intelligent software platforms. Welding remains a skill or craft to be digitised (DT5) on low volume, unstructured parts presenting cost barriers. i3D have developed collaborative augmented welding systems controlled by 3D stereo cameras & AI enabled SGM inspection software (aerospace component remanufacturing) for RT defect detection & characterisation. Retraining our bespoke ML data sets for welding defect inspection on the Pilotʼs chosen material (& illumination) is enhanced by DL techniques & generative AI defect simulation. i3Ds vision software automatically produces robot control for collaborative welding & therefore possible adhesive dispensing toolpaths for irregular shapes (DT4) thus enabling QA inspection solutions (DT7). I3D will present these outputs as HMI interface dashboards enabling human supervision to decide on which data or control function is the “best fit” to proceed with manufacturing, enabling ease of integration with existing factory operations. 

Close

SINDA

SINDA focuses on overcoming the fragmentation of data and lack of interoperability among façade design, simulation, and lifecycle management tools. Current façade workflows rely on isolated repositories and non-standardized formats that hinder collaboration, traceability, and efficient use of digital tools. SINDA will provide an innovative, scalable, and standardized data backbone to integrate heterogeneous sources and harmonizes design, simulation, and operational data into a single interoperable repository. The platform will connect with DT1 (AI-enhanced parametric design system) and DT2 (automated façade simulation) to consolidate design parameters, environmental conditions, and simulation outputs, and integrate with DT3 (extended digital twin) acting as the synchronization hub in order to allow real-time data exchange and lifecycle traceability. SINDA novelty lies in its modular, FIWARE-based architecture, adopting open standards (IFC, OPC-UA, ROS, ISO 19650, SAREF) and security-by-design mechanisms (TLS, Keycloak, XACML) to ensure interoperability, scalability, and secure data sharing.