Call2 Monitoring dimRob

Public Summary Month 8/2012

Fri, 03 Aug 2012 15:25:42 GMT

In the final phase of the experiment, objectives of the research were compiled in a final demonstration and in addition, the concept of using the fabrication unit as a digital workbench that is able to accomplish various necessary tasks in a fabrication sequence was introduced in order to show that the unit was able to react on the changing conditions a “workspace”, the building site.

After handling each task in the sequence separately, they were all put together in a single fabrication chain that consisted of: (1) Measurement of the inclination of a workspace and orienting the work tool (gripper) accordingly by transferring the angle data to the robot controller, (2) scanning the workspace (inclined frame) and importing it as 2D-lines to the design software (3) adapting the algorithmic design of a wall structure to be built in that workspace according to the scanned and recognized changes in the dimensions, inclination, etc., and (4) robotically cutting and placing adapted and custom shaped building elements of this algorithmically designed wall structure.

Public Summary Month 6/2012

Sun, 10 Jun 2012 15:55:35 GMT

Regarding the preparations of the fourth (final) demonstrator, these two months, mainly the focus was to handle the scanning of the steel frame (which is going to be the workspace in the demonstrator), finding the right orientation of the gripper regarding the inclination on the workspace and the cutting tasks separately. All these tasks were successfully accomplished with small mock up demos, and prepared to be put together in the main (fourth) demonstrator in the following weeks. In addition to involving all 3 of the experiment objectives; in this fourth demonstrator, the intention is to use the robotic unit as a digital workbench as if on a construction site.

Public Summary Month 4/2012

Thu, 05 Apr 2012 09:36:45 GMT

Following the pioneering demonstration in late January 2012, in which a modular 8-meter wall was assembled in 1:1 scale by 13 consecutive self-positioning of the mobile fabrication unit in an unknown environment (the parking space of ETH Zurich), this month, mainly the focus was to document and publish the progress achieved so far, in addition to the preparations done for the fourth demonstration. Next demonstration will involve all 3 of the objectives; being mobility, man-machine interaction and handling of material tolerances; also with the intention being to alter the local reference system into a global one in several terms. Also, the main progress involving Task 3 is concluded, still, this progress will be used in the 4th demonstrator.

Public Summary Month 2/2012

Mon, 20 Feb 2012 15:21:45 GMT

Complementing to the prior experimentations in previous months on man-machine interaction and handling of material tolerances, current stage of the project focuses on the proof of the concept of mobility.

A system is developed for self-positioning of the mobile unit through the 3D scanning of a local reference. In the demonstration, an 8-meter long modular wall is assembled by 13 consecutive self-positioning of the mobile fabrication unit in an unknown environment (Fig 1). Next stage of the project involves an integration of the 3 objectives; being mobility, man-machine interaction and handling of material tolerances.

The demonstration that took place in the last month was the pioneering experimentation on putting the mobile fabrication unit in its context – on the construction site. The context is chosen as the parking space of ETH Zurich with a ceiling height of 2.50 m. It is the pioneering demonstration for building a large building element on-site.

Initial Tests on Relocation (proof of concept with 3 consecutive moves)

Prior to the fabrication of the fragile structure (the wall) in the parking space, a small scale demonstration took place for proving the concept of mobility according to a local reference system.

In this system, the 3D scanner which was implemented on the robot is used to self-position the mobile unit. The scanning mechanism finds the center point of the metal disk and this point is set as the origin of the plane (coordinate system) that is defined in the CAD model for each new position.

Real-World Conditions

The data from the CAD model of the structure is imported to the robot controller using the system developed on the software side. The wall, the fragile structure, is designed and partitioned according to the simulation studies done with offline programming for finding out the most feasible positioning of the fabrication unit.

The whole design complexity of the structure arises from the superposition of algorithmic rules with cannot be applied easily in a manual fabrication process. Several aspects of the design is driven by the robotic fabrication method chosen.

Fig. 1: The fragile structure: A modular, robotically fabricated 8-meter long wall.

Public Summary Month 12/2011

Mon, 05 Dec 2011 09:53:18 GMT

The tests and framework studies that have been done in the last two months, belong partially to Task 2 (Positioning/Relocation), partially to Task 3 (Site Tolerance) and partially to Task 4 (Material Tolerances). Current stage of the project requires an integral progress on all of these 3 tasks and it aims proving the steps of the hypothesis developed for the final demonstration. Based on the theoretical study/literature survey on robots in construction sites two hypothetical scenarios were developed for two different cases, and a hypothetical framework (Fig 1) was set for these two scenarios. Both of the scenarios involve building a modular wall with the robot on the construction site either with reference to existing elements in the workspace or in an undefined workspace.

Fig. 1: Hypothetical framework set for the two scenarios developed

Setting a hypothetical framework clarified all the steps to be taken from start to end for the developed scenarios. Depending on the scenario chosen, hardware, method of data handling, utility and controlling mechanism changes.

Description of the Hypothetical Framework

As it can be seen in Fig 1, framework consists of the following stages:

Transportation of robot to construction site,
Transportation of material to construction site,
Preparing the set-up (power),
Positioning
Calibration (3-point calibration)
Object recognition / measuring
Feedback of tolerances
Calculation of solutions
Data communication
Re-positioning of robot
Re-positioning of material
Starting to build
Monitoring

Tests on Self-Orientation

In this test, 3D scanner is programmed to measure 3 taught points from the tilted wooden board and by drawing a plane through these points, it finds the right orientation for the gripper.

Fig. 2: 3D scanner is programmed to measure 3 taught points from the tilted wooden board and by drawing a plane through these points, it finds the right orientation for the gripper

Tests on Self-Positioning

The 3D scanner which was implemented on the robot was also used to test self-positioning with the help of a metal disk whose central point was used as a reference marker (satellite). With the satellite, robot will be able to move several times, re-positioning itself according to the measured coordinates of the center point of the satellite.

Fig. 3: A 3D scanner attached to the gripper of the robot finds/measures the center point of the steel disk (satellite) and gripper moves to that measured point accordingly

Tests on Material Gripping

Tests done on material gripping cover trying out the material size that is being planned to be used in the final demonstration, and building a material dispenser out of aluminum profiles integrated to the robot. This material dispenser is adaptable to different material sizes and attached to the robot.

Public Summary Month 10/2011

Mon, 03 Oct 2011 13:21:48 GMT

In the last two months we have been working on Task 2 (Positioning/Relocation) and Task 3 (Site Tolerance). Task 3 deals with the challenges of dimensional tolerances existing on construction sites. The objectives are, firstly to measure the real world data and secondly to develop a software prototype which maps the fabrication data to the acquired real world data. For demonstrating the human-robot cooperation we have prepared an installation for ETH Scientifica exhibition 2011 in Zürich.

http://www.dfab.arch.ethz.ch/web/e/forschung/216.html

For Scientifica 2011, the focus has been on robotic fabrication strategies in architecture in the scale of 1:1. The task consists of building an architecturally complex brick wall. In the demonstration robot unit employs an 3D scanning system which enables it to orient itself and at the same time to process the information gained from different materials and from the surrounding environment.

Fig. 1: A 3D scanner recognises the move of the hand, and by this way the hand move is recorded to build accordingly.

Fig. 2: Building according to the recorded hand movements (along the drawn line)

Fig 3: After the robot grasps the move of the hand, it builds the new position of the wall in an infinite loop accordingly.

Since the conceptual phase is now almost over, a theoretical research on the real case scenarios of the construction sites is needed to develop the concept into a real case prototypical demonstration, with the feedback coming from the current state of art of construction sites. Therefore, besides the demonstration we also went through a research on the past and the current state of art of robotics on construction sites. Some current robotic construction techniques involve brick laying, welding, cold folding, crack detection, surveying through solid walls, fitting equipment into inner walls, floor finishing, panel fixing, exterior wall painting, concrete distribution, application of bonding material, smoothing concrete, etc.

Public Summary Month 8/2011

Thu, 25 Aug 2011 19:35:16 GMT

In view of a worldwide shortage of energy resources and an increasing concern about climate change, the research project dimRob focuses on a radically new approach for tomorrow’s building reality. dimRob studies the manifold amalgamation of architectural design and robotic fabrication and researches on the invention of novel technological solutions within the context of design, construction and fabrication. Through introducing mobile accessibility, cognitive skills and highly innovative scanning technology dimRob fosters the next logical step in robotic fabrication and addresses multiple potentials in making the robot operational.

Key to this approach is the underlying notion that robotic fabrication is traditionally embedded within the industrial environment where fixed positioning and constant conditions determine the robot’s life. However, extracted from this safe context in order to directly use it on the construction site, multiple issues arise – on one hand, for this purpose the robot has to be mobile and robust; it has to recognize its own position and to assemble new building elements within existing construction and material tolerances; on the other, it has to operate on human control and to react on visual instructions. This forms the basis for an innovative research approach and forms the empirical methodology of this project. In this regard, the first two phases of scientific investigation can be summarized as the following:

During the first work package of dimRob, not only the experiment set up could be successfully implemented and already used for a larger exhibition project (2011 Fabricate Conference, London) but also the additional hardware equipment could be fully integrated. Together with the industrial partner Bachmann Engineering AG a compact mobile track system with integrated raising jacks was developed and formed the basis for mounting the ABB IRB 4600 (2.55) industrial-robot on it. The ABB robot not only provides with a wide operational range but also offers low weight. Moreover, in order to use the robot unit on a construction site, the boundary dimensions were limited to the size of a standardized door frame. This also allowed for a compact transport unit. To further advance the operationability, a diesel engine was integrated for the transport of the robot unit.

Further, the robot unit was equipped with two vacuum grippers in order to use it for particular (additive) construction purposes. This enabled taking single bricks at high speed and to position them with high precision throughout the operational building space of the robot unit. In order to have a compact mobile unit, the vacuum pump for the gripper was installed on the back of the robot. In addition, a powerful laser measurement system (Sick LMS 500) was installed on the robot’s arm. It guarantees the detection of tolerances on the construction site and emphasizes further research in this field. Next to the hardware setup, also the development of software played a major role in this phase of development of dimRob. For the buildup of individually articulated masonry, a plug in for an existing CAD software was programmed. This enabled to export data directly from the design to the programming language of the robot. In the next phase of development, this tool will be further refined and adapted to more efficient workflows and broader applications.

The second phase of investigation is determined by developing a cognitive strategy for self-positioning the robot. As the unit is planned to be used on construction site it has to operate within partly unknown contexts and has to freely orientate in three-dimensional space as well as to recognize obstacles or human workers. This is enabled by the already implanted scanning technology and will be further developed by programming spatial feedback operations that compare collected data to already existing data of the site. Here, the 2011 Fabricate Conference in London provided with a first useful opportunity to test this scenario and allowed for constructing an additive assembly of individual wooden building-elements based on an intelligent feedback strategy.