top    Laboratory of Artificial Intelligence for Design

creating smarter
design tools
the quest for
artificial imagination
first online 3D
sketch interpreter
now machines
think spatially
free 3D CAD
for experiments

a project of Lai4d Systems  

The design of 2D or 3D geometries is involved in most of science, art and engineering activities. Modern design tools are powerful and boost the productivity of designers, but require a lot of training, effort and time to achieve a good understanding and an efficient exploitation. LAI4D is a non-profit R&D project whose aim is to develop an artificial intelligence able to emulate cognitive functions regarding spatial imagination and other capabilities. This technology will help improve the communication between designers and design tools as well as accelerate the engineering process.

video The practical implementation of the research has been conceived as a dual web application that can work as a 3D viewer widget or as a free light 3D CAD tool providing the adequate environment for the project. This CAD tool incorporates a special design assistant capable of extracting conceptual geometries from pictures or sketches provided by the user as input thanks to innovative AI algorithms.

Additionally LAI4D tries to reduce the inherent complexity of professional design tools which, despite being suitable for experienced users, are almost unreachable for other people not trained in the usage of CAD systems and only in need of an occasional use. The selected implementation approach not only allows the users an easy access to the tool, but is also an excellent mean to build a community of designers that will provide the necessary feedback for the system in order to make it bigger and smarter. Start creating 3D models with the LAI4D designer.


LAI4D source editor The LAI4D widget is a minimalist 3D viewer and designer for the web, all in one, and is used as a test-bed for the research. This widget can be easily embedded in other web pages in the form of an IFRAME element whose URL indicates the drawing source. The viewer allows to render and explore 3D drawings in web pages without the need of plugins even if the device is not WebGL enabled. The designer is the free drawing tool for the people without the time to learn the usage of a professional CAD application and that need results in minutes. It allows to create 3D drawings by writing the description of the geometries using a really simple language.

LAI4D viewerAs opposed to most modern applications, direct source edition as plain text takes again relevance in LAI4D because it is straightforward for occasional and inexperienced users requiring a minimum learning effort. At the same time, the created drawings can be shared through the Internet uploading them to the LAI4D server from the own designer widget. And if a more advanced usage is required, a rich command and graphic interface is also available.

Although LAI4D has not been conceived as a professional CAD tool, it can handle any kind of design since it offers the following elements:
  • Polygonal lines and polyhedrons.
  • Spheres, cones and revolved surfaces.
  • Curved lines, pipes, organic surfaces and meshes.
  • Signals with texts and images.
  • Algorithmic shapes.
  • Molecular structures.
Additionally LAI4D offers other interesting features such as:
  • Lai4dCL which is one of the simplest text based drawing formats, and much more comfortable than XML based formats.
  • The inheritance strategy that allows to easily define an entity or part of an entity from another one.
  • The use of sub-drawings logically contained in an entity making the drawing to work much the same as a navigable tree of folders.
  • The possibility of attaching hyperlinks or custom user data to any entity of the drawing.
  • The algorithmic generation of entities requiring only JavaScript knowledge from the user.
  • A free online storage service that allows the immediate sharing of the created drawings through the Internet in case the user doesn't have his own web space.
  • The control of entities, view configuration or behavior options through URL instructions.
  • Import and export capabilities for other drawing formats.
  • A 3D sketch interpreter that uses AI to conceptualize hand-made drawings.

LAI4D designer UI All graphic functions of LAI4D are based on the capabilities of the canvas HTML element. Although major web browsers are WebGL enabled, in case of need LAI4D can work both with WebGL and canvas 2D rendering contexts. Obviously the performance when using canvas 2D is less than that when using WebGL (accelerated graphics hardware) but, at least, it allows to handle 3D geometries in any device.

LAI4D can be used not only online but also offline. The LAI4D web page downloaded from the Internet can be saved to a local drive of the client's machine using the standard "Save as" function of the web browser normally located in the "File" menu. This will allow the user to work offline regardless of whether the LAI4D page has been cached or not by the browser. This is particularly necessary if the user wants to create a distributable HTML document containing embedded LAI4D drawings as IFRAME elements that can be displayed both online from the Internet or offline from the local drive of a computer.

The user interface is structured in order to allow an easy access to the functions. The most common functions are grouped at the corners of the viewport while the rest of functions and configurations are organized in a tree view control.

The detailed functionalities of LAI4D can be found in the reference manual at the documentation section.


We taught the machines to understand our speech. We taught the machines to understand our writing.
And now we are teaching the machines to understand our spatial ideas

video Designers spend an important part of their work time transferring decisions to the design tool. This is normally done by means of commands configured with coordinates or other parameters. The design assistant available in the designer widget is an experimental tool intended to provide a new way of communication between the user and the design application. Its goal could be explained as the capability of understanding the user ideas. When the assistant analyzes a user's input, what it is essentially doing is trying to understand that input in order to generate a canonical or formal response ready to be exploited for design purposes. The tool can be basically used in two ways:
  • Entering a text string in order to be parsed obtaining the corresponding formal data structure, or in order to find help about a topic.
  • Making a sketch or uploading a picture in order to obtain the corresponding conceptual geometry after a heuristic interpretation process.

box conceptualizationThe sketch interpreter of the design assistant is the face of the conceptualizer program JAIC. The goal of this program is to "imagine" the conceptual geometry represented by an imprecise raster image in a probabilistic way regardless of whether it is a recognizable shape or not. The sample image shows a set of irregular strokes at the left side, and it is what it is, a sketch with irregular strokes. Despite their imprecision, a person could also say that the strokes represent a box, like the one at the right side. The necessary process to reach to the conclusion that those irregular strokes are representing a perspective view of a box is a conceptualization process. While a box is a well-known shape, the conceptualization can also be extended to the comprehension of non-recognizable and never-seen-before geometries when such geometries have some kind of sense.

JAIC only uses recognition for the OCR function. Convolutional neural networks are very good for finding similarities between an input and a database of patterns, like in the OCR, but this approach is not enough to solve the problem of conceptualizing generic images. Although humans count with biological 3D sensors like stereographic view and ocular focus, much of the image interpretation is achieved through conceptualization. Indeed when a human watches TV or plays a computer simulation game or remotely pilots a vehicle through a screen, is the conceptual vision the only mechanism in charge of building the 3D scenes on his mind because no other 3D information is actually available. The LAI4D research is an indirect approach to the understanding of the human conceptual vision which probably is one of the key pieces for an enhanced motor dexterity of autonomous robots and higher cognitive capabilities like artificial imagination.

Sketch interpreter UIJAIC is progressively implementing the necessary knowledge to try to understand the 2D or 3D conceptual geometry represented by sketch images under certain conditions. When the analysis is successful the program generates the corresponding canonical geometric entities offering the possibility of adding them to the drawing. Although the current capabilities of JAIC are still very limited its potential is great.

Traditionally humans share complex ideas through plans, sketches or other graphic means because they are more efficient for transmitting the huge amount of information needed to describe such ideas. Unfortunately this strategy does not work so efficiently when the recipient is a machine since it usually demands the information in a canonical format. But, what if the machine were able to understand the user's ideas? That is the path LAI4D is trying to open by developing JAIC.

This technology is absolutely experimental and is at a very early stage of development. A big effort has been done for implementing the first versions as a client side application offering useful offline capabilities, but this cannot be a long term strategy due to the limited power of personal computers. In the future most of its functionality will run on remote servers better fitted for executing the heavy queries and brute force algorithms involved in artificial intelligence tasks. See the section "Working with sketches" of the Reference manual for more details.

Although the research carried out by LAI4D is quite singular due to the fact of not being based on conventional deep learning, there are other research projects or organizations pursuing similar goals. Next it is offered a list of interesting links that can be classified under the concepts of sketch recognition, 3D reconstruction or 3D shape retrieval, as well as others with a more general AI scope:


Click on the image links to open the examples in a new window. Some of them may take several seconds to render. All details about the entities used to compose these examples can be found in the reference manual at the documentation section.

Paper airplane
A very simple geometry representing a paper airplane
It is composed of a single polynet with 4 faces.
No styles are used so default colors are applied.
This example composed of a single "curve" entity representing a human face profile demonstrates the versatility of this type of entity. If two consecutive vertices are equal then the corresponding segment evaluates to a single point. If two consecutive vertices are equal and the next two consecutive vertices are also equal then a straight segment is generated between the two pairs of vertices. Thanks to this the "curve" entity can represent linear geometries with corners mixing straight and curved segments.
In this case all the vertices of the curve are placed at the same plane so it is a 2D geometry, but the "curve" entity also accepts 3D vertices. If the vertices of a regular hexagon are used as vertices of a closed "curve" entity, then the obtained geometry looks much like a circle.
The example is composed of a "pipe" entity. This type of entities are defined by a profile and a guideline, and both can be either open or closed. Organic surfaces may take several seconds to render.
Profiles or guidelines can mix curved and straight segments and can have corners since they are defined through the same algorithm used for the "curve" entity. A profile can be defined by a set of 3D points so does not need to be planar. The profile is assumed to be referenced to the origin {0,0,0} and to the orientation defined by the Z axis. The surface is built by translating copies of the profile from the origin (reference point of the profile) to the points of the guideline and, at the same time, reorienting the profile copy making the reference Z axis tangent to the guideline.
Extruded surfaces can be generated using a straight segment as guideline.
This design is an example of the use of "patch" entities. There are only two original surfaces: one patch for the delta wing and other patch for the air intake. The rest of surfaces are obtained by means of symmetries from the two previous ones. Note how the border lines of the patches can be either straight or curved lines. Organic surfaces may take several seconds to render. Some lines have been used to decorate the drone.
There is a hidden entity (a tri-axis only used as reference and placed at the zone of the outlet nozzle) with an attached sub-drawing containing auxiliary entities used during the design.
Curve inherited from line
This is an illustrative example of the use of inheritance. Inherited entities must have a "name" member while heir entities must have an "inherit" member indicating the name of the entity from which to inherit. All members of the inherited entity will be "copied" to the heir entity with the exception of those already present in the heir entity which will be preserved.
The drawing contains a "line" entity whose name is "a" defined by 6 vertices each of which has been assigned a color. The drawing also contains a "curve" entity whose definition is inherited from that of the entity named "a", the "line" entity, so it passes through the same vertices. Note how the line is open while the curve is closed because the curve has a member "closed" set to a non-zero value.
Tetrahedron with signals
This example is composed of a single polynet that incorporates the four available types of signals. All the signals are anchored to the same vertex. The polynet has a member "link" (attaches a link) that points to the drawing itself so, when navigated, the same drawing is opened in a new window.
The dot signal is placed directly over the anchored point. The polygon signal is used as a leading line joining the vertex and the text. The text signal has an offset with respect the vertex. The source of the image signal has been codified in Base64 instead of using its path so it is not an external resource but an internal one.
Spiral screw
This is an example of a "revolved" entity. By default this type of entity builds a surface of revolution with the given profile around the Z axis, but it can optionally be a screw like surface providing the "advance" member, as it is done in this example.
Profiles can mix curved and straight segments and can have corners since they are defined through the same algorithm used for the "curve" entity. A profile can be defined by a set of 3D points so it does not need to be planar. In this case the profile curve represents a 3D spiral.
Cones and cylinders can be generated using a straight segment as profile.
Organic surfaces may take several seconds to render.
This example represents a simplified distribution of the walls of a house. Each wall is composed of a single "face" entity, even those with windows. The trick to build a face with holes is to consider the borders of the holes as borders of the face. In theory, the vertices of a face should be in the same plane but LAI4D does not perform any checking. It simply projects the vertices and paints the corresponding polygon which may seem strange if that set of vertices is not well constructed. The black lines are auxiliary and can be hidden, they are not part of the faces. One of the walls defines the wall color and the other ones inherit it. The selected color is semitransparent.
This drawing includes a view definition and a configuration for avoiding the camera inclination.
Algorithmic snail
The geometry represents a multicolor snail shell. It has been obtained through a program that implements the generator algorithm using as arguments the section's shape and the number of revolutions. The result of the algorithm is a drawing source file containing a single polynet composed of colored faces. This source is not very long but the execution of the algorithm can take several seconds.
It is impossible for a CAD system to provide a set of predefined entities suitable for all imaginable uses. The usage of programming languages that allow the generation of customized source files solves this limitation. For helping in this purpose LAI4D offers the special entity type "program".
Random city
A "program" entity generates a set of "box" entities that looks like a city. The user can provide the ground width, the ground length, the maximum box height and the number of boxes through the "variable" member. The greater the number of boxes is, the longer the time to render will be. The algorithm randomly decides box sizes, positions and colors trying to set greater box heights at the center of the ground area. Each time the entity is rendered a new distribution is obtained due to the usage of random functions.
As any other entity, the result of a "program" entity can be converted into polynets using the corresponding function in the LAI4D designer.
This drawing is an example of how "molecule" entities can handle a great number of atoms efficiently from a rendering point of view. It corresponds to a fantasy molecular structure built from four dodecahedrons and using four atom types randomly arranged.
The atoms painted by a "molecule" entity are not real spheres but signals composed of images and circular dots which notably reduces the graphics load. In LAI4D molecules can also show the atom symbols and each atom can have either default color and radius or customized ones. The color of the atom links is also configurable. The background color has been made black in order to obtain a better contrast.
In this case the link opens an HTML page with the LAI4D widget embedded in an IFRAME, instead of showing directly the widget's page. The parent page defines a background image which can be seen because the widget is transparent. The model represents a fancy fish composed of colored patches, spheres, filled curves and symmetries. Inheritance is also used.
When a linear entity like a "curve" defines the member "fill" with a non-zero value, it is represented as a face. This strategy is used for some of the fish's fins.
Multi-view device
The link opens an HTML page showing 4 views of a device. The HTML page contains a table with 4 cells. Each cell contains a LAI4D widget embedded in an IFRAME with a customized URL.
The URL start-up instructions of all the IFRAME elements point to the same file which defines the device's model. In return, the different view configurations used for each widget are passed using the member "view" of the start-up instructions.
The geometry of this spider with ant head has been imported from a .PLY file found at: (Greg Turk)
Once imported, a color and a curved face effect has been applied.
The currently supported import and export capabilities are covered by the functions under the "File" menu of the LAI4D designer.
The geometry of this trumpet has been imported from a .OBJ file found at:
Once imported, a color and a curved face effect has been applied.
The currently supported import and export capabilities are covered by the functions under the "File" menu of the LAI4D designer.
Interpolated tetrahedrons
Two similar "program" entities generate artistic representations of tetrahedrons using the function "AppLai4d.GInterpolate".
The algorithms draw lines joining the segments of the tetrahedrons using simple linear interpolation to get the line points and point colors. One of the algorithms distribute the lines regularly while the other one makes it randomly.
Fractal fantasy 1
A "program" entity that generates a 3D fractal geometry composed of regular tetrahedrons.
In each iteration a new tetrahedron is placed at each free face of the tetrahedrons created in the previous iteration. The three vertices of the contact face of a new tetrahedron are obtained by linear interpolation from each of the three edges of the contact face of the old tetrahedron. An optional list of colors can be distributed among the resulting vertices.
The number of iterations, the interpolation parameter and the optional list of colors can be configured through the "variable" member.
This entity is another example of the usage of several functions provided by the JS object "AppLai4d" for designing algorithmic entities.
Fractal fantasy 2
Another example of "program" entity generating a 3D fractal geometry composed of regular tetrahedrons.
In each iteration a new tetrahedron is placed at certain distance of each vertex of the tetrahedrons created in the previous iteration. The new tetrahedrons are scaled and rotated. An optional list of colors can be distributed among the resulting vertices.
The number of iterations, the distance factor, the scale factor, the angle and the optional list of colors can be configured through the "variable" member.
In this case the functions provided by the JS object "AppLai4d" have been used in the optimized way minimizing the number of arrays to be created in each operation.
Simple 3D math
This example shows how easy is to obtain the 3D representation of a "z=F(x,y)" function using a "program" entity.
In this case the algorithm returns a "mesh" entity whose "table" member is convenient for arranging the data generated by this kind of functions.
A set of vertex colors are calculated for artistic purposes.
Configurable spindle
The link opens an HTML page showing an IFRAME at the left and a web form at the right. The form allows to configure several parameters which define the geometry of a mechanic spindle. The button "REGEN" takes all parameters and generates a new URL for the IFRAME in order to paint the expected spindle.
The URL start-up instructions of the IFRAME always points to the same file which contains a very simple "program" entity. Its algorithm needs a set of variables also passed through the start-up instructions. This is a good example of how the "append" member of the start-up instructions can be combined with inheritance to get a drawing configurable via URL.
Knot mesh
This shape is a "mesh" entity. A mesh is a simple tabulated surface based on an array of vertices. Thanks to the optional member "smooth" the surface can be curve instead of polyhedral. Furthermore the surface can be defined as closed by the rows or the columns. This entity type is convenient for building irregular surfaces from arrays of points or organic surfaces based on frames.
The table of vertices of this example have been extracted from a "pipe" entity defined by a profile and a guideline. This original pipe entity, with almost the same geometry, can be found navigating into the children of the mesh.
Illusion ring
The shown geometry is built up with a single tape obtained from a "mesh" entity.
It could be explained as a closed pipe in which the section is a triangle that rotates 120° along the guideline over its own plane. The tape takes 3 complete rounds ending at the same point where it begins so it is closed.
Navigate into the children of the entity to see the auxiliary constructions used to design it.
Bitten apple
This model has been obtained as the result of intersecting two revolved entities, one representing an apple and the other one a bit.
The stem is a pipe entity.
Navigate into the children of the stem to see the auxiliary constructions used to design it.
Organic surfaces may take several seconds to render.
Terrain generator
A "program" entity is designed to generate a terrain piece with colored mountains and a water plane using a very simple pseudo-random algorithm.
The "variable" member allows to decide the ground side, the maximum height, the density of the mesh, the possibility of making it smooth and the color list that will be interpolated according to the heights.
Always the drawing is regenerated a new terrain is obtained due to the usage of random functions. Open it in design mode in order to modify the "variable" member.


Beginner's tutorial: This tutorial is the recommended introduction for all persons new to LAI4D. It is intended to teach the basics of design in 20 minutes. It shows step by step how to create a simple 3D geometry from the idea up to the publishing of the drawing on the Internet using the easiest tools. Thanks to this exercise the user will understand the working philosophy of LAI4D, will be able to generate polyhedral surfaces and polygonal lines with colors, and will learn to share designs online. The created geometry can be inspected in the link: cubicle_sample.

How to design a 3D model with LAI4D: This video-tutorial covers the designing of a 3D model using different strategies as well as its later saving to share it online. It is based on the Beginner's tutorial of the LAI4D designer but extended in order to show also how the user can work with the menu of commands or the sketch interpreter.

UI quick help: This help document summarizes the functions available through the user's interface of the viewer widget.

Reference manual: This document is the most complete guide to the functions of the LAI4D widget but it is not a tutorial. It is indicated for people willing to make a deeper usage of the tool. The following topics are covered by the reference manual:
  • Lai4dCL. The language used for describing the source of LAI4D drawings and other data structures. Syntax, inheritance strategy, entity model, sub-drawings...
  • The standard and the extended entity sets. It explains the structure and properties of the different entity types available to compose LAI4D drawings. Polynets, lines, surfaces, transformations, algorithmic entities, molecules, view definitions...
  • User interface. This section explains the usage of the functions available through the LAI4D's UI. View control, file management, configurable options, edition and query operations, design assistant...
  • Publishing LAI4D drawings. Sharing of drawings, configuring start-up instructions through the URL, offline usage...
  • LAI4D architecture. Additional information for achieving a better understanding of the LAI4D functions and limitations.

Find more interesting tutorials and articles in the community.


+ What can I do with LAI4D?
LAI4D widgets allow to render, explore and edit 3D drawings in web pages without the need of plugins. The designs can contain lines, surfaces, texts, images... The drawings can be edited through graphic commands, as plain text using an intuitive language, or using the design assistant. Additionally LAI4D implements an amazingly simple strategy for generating algorithmic entities. Furthermore, the created drawings can be immediately shared through the Internet uploading them to the LAI4D server from the own widget.
+ Where do I start?
Visit the documentation section. There you will find tutorials and reference manuals. The "Beginner's tutorial" is designed to learn in 20 minutes how to create a basic 3D design from the idea up to the publishing of the drawing on the Internet.
+ Is LAI4D free?
Yes. The standard LAI4D widget, both viewer and designer, can be used for free as is without modifications. It can even be downloaded and used offline if necessary. The function for storing drawings in the LAI4D server is also free and no registration is required.
+ Is LAI4D open source?
Not currently, although this option is not discarded for the LAI4D project in the future. LAI4D has been built from scratch over the Iquix framework. Neither the rendering engine nor the design application nor the intelligent assistant use external libraries or components. Iquix is a proprietary JavaScript library that ranges 3D vector graphics, computational geometry, physical simulation, artificial intelligence, multi-thread execution as well as other more conventional functionalities.
+ Could the LAI4D widget be customized?
Lai4d Systems can customize the LAI4D widget changing aspect, functions and behavior according to the client requirements. Contact Lai4d Systems for more information about this possibility.
+ Is there any API for exploiting the LAI4D functions programmatically?
The LAI4D widget allows the passage of start-up instructions through the URL, see the "Widget initialization" section of the Reference manual for more details. Although it is not an API, LAI4D implements an entity type called "program" that allows the algorithmic generation of graphic entities through JavaScript programming. A public JavaScript API could be released in future versions of LAI4D.
Lai4d Systems can provide consulting and development services for customizing or integrating the LAI4D functionality in other software.
+ Can the LAI4D widget work with other drawing formats?
The currently supported import and export capabilities are covered by the functions under the "File" menu of the LAI4D designer.
+ How can I send error reports, suggestions or queries to the LAI4D team?
User participation is centralized through the LAI4D's community. Error reports, suggestions or general queries can be posted there where members of the LAI4D team or other users will reply a.s.a.p.
+ How can I embed a LAI4D drawing in my web page?
Once you have a design ready for publishing you will only need to save it to the free online LAI4D storage service, and then to insert in your page the right HTML code for the IFRAME element that will show the drawing. In order to save the design use the command "Save server file" under the "File" menu. The simplest way to obtain the IFRAME code is to use the "Share" button at the top-left corner of the widget.
Alternatively you can also store a drawing file in your own web space and instruct the widget to open it directly from there using its absolute URL. A more detailed explanation can be found in the section "Publishing LAI4D drawings" of the Reference manual.
+ What is a polynet?
It is the intrinsic graphic entity of LAI4D. A polynet defines a set of vertices, a set of faces, a set of lines and other properties for colors or line widths for example. Any other type of graphic entity is to be converted into a polynet or a set of polynets during the interpretation process before it can be rendered.
+ How can I include texts and images in a drawing?
This is done through "signals" that must be integrated in a polynet entity. Signals are 2D graphic entities that are always painted parallel to the projection plane regardless of the camera orientation. They can be images, texts, circular dots and polygons. Each signal is anchored to a vertex of the polynet so it has a 3D position.
+ Why is so easy to use algorithmic entities?
What a "program" entity does is essentially to execute a JS algorithm that returns a standard LAI4D entity that can then be rendered in the usual way. The Lai4dCL syntax is so simple that the source of a LAI4D entity can be immediately represented as a JS array or vice versa. Thanks to this characteristic the result of the "program" entity can directly be a JS array instead of a Lai4dCL structure so there is no need to deal with specific APIs to generate algorithmic entities. This advantage combined with the fact the LAI4D entities have a very intuitive format make "program" entities really easy to use.
+ How does the LAI4D widget exploit artificial intelligence?
The use of AI is centered in the design assistant. Its sketch interpreter is an experimental tool able to extract the conceptual 3D geometry represented by the given sketch (hand-drawn or loaded from an image file). The result can be later edited in the model space as a normal entity.
+ Which is the difference between shape recognition and shape conceptualization?
A shape recognition process attempts to find a match for the analyzed image (or other input) within a shape descriptions data base returning as result the found record if any. However, a shape conceptualization process attempts to "imagine" the shape represented by the analyzed image regardless of whether it is a recognizable shape or an abstract one returning the imagined shape's geometry as result.
+ How can I get involved in the project?
The easiest way to get involved is to become an expert user in order to participate in the community helping other users, writing tutorials and providing feedback to the LAI4D development team.


The community is a space intended for cooperation between LAI4D users and developers. All users participating in the community are helping the project in one way or another. To answer questions, to report a bug, to write a tutorial, to share a useful algorithmic entity or to discuss an enhancement proposal, all are welcome contributions. After being evaluated by the LAI4D development team, this feedback will become into software and documentation improvements. By participating in this community you also accept the terms and conditions expressed at the bottom of this page.
Remember that you can also contact to the development team at If you want to contribute with tutorials or interesting articles send us an email.

LAI4D's board at
LAI4D's blog at Medium
Topic: Tutorials
LAI4D's YouTube channel

LAI4D in the media:


The estimation of hidden faces has been improved.
The brightness calculation scheme has been upgraded including a new factor "focus reflection factor" indicating the proportion of reflected directional light that is focused instead of diffuse.
A "copy to clipboard" button has been included within the information elements of the commands: "Point coordinates", "Measure distance", "Measure angle", "Measure polygon", "Physic properties".
The command "Create extruded surface" has been replaced by the command "Create pipe". An extruded surface can be created entering a straight segment as guideline of the pipe.
Other minor improvements.
+ Older versions
The manual input of polar coordinates has been enabled through the new button (<) in the input area.
The command "Transform entities" has been enhanced making only the first pair of points mandatory. Depending on the number of pairs of points provided the operation can be: simple translation, translation + rotation + elongation along a line, translation + plane alignment +  planar deformation, translation + complete transformation.
The prototype of the entity "transformation" has been modified in order to resemble the behavior of the command "Transform entities". Now the transformation is defined through the member "pairs" which can hold up to four pairs of points. The old behavior using the members "matrix", "origin" and "move" was of little use since any entity can declare the members "transform" and "translate". This new prototype can be particularly useful for creating algorithmic entities.
The function for pausing the command has been also implemented for the selection of entities. If used then the current command is paused and the user can change the camera configuration in the usual way.
In the sketch interpreter, the interpretation of holes sketched in faces has been improved. In order to be correctly handled, the loops corresponding to the holes must be connected to the external contour of the face. The resizing of faces with holes and assigned measures has been improved. See the section "Working with sketches" of the Reference manual for more details.
The "dynamic zoom" navigation has been implemented. Each mouse wheel event produces a combination of a zoom increment plus a camera movement. To use the dynamic zoom is a comfortable way of exploring a wide scenario without the need of calling several times to different camera configuration buttons.
Now the sketch interpreter allows the loading of pictures into the sketch area through drag & drop operations.
Now the designer widget allows the loading of drawing files into the canvas area through drag & drop operations.
The text resolution in WebGL contexts has been increased.
The performance of several CAD commands has been improved.
Assignment of measures to segments improved.
The configuration "Inverse depth interpretation" has been added to the sub-menu "Design assistant options". By checking this option the design assistant changes the depth interpretation criterion when generating 3D hypothesis.
Now the design assistant tries to find the evident reference system of the 3D hypotheses when passing them to the model space. If found then the hypotheses are reoriented before placing the new geometry at the model space.
The new command "Cut by plane" calculates the line and surface fragments resulting from the intersection of a set of entities and a plane defined by 3 points.
Several improvements in the sketch interpreter.
Imagination of hidden faces improved. Now this function can work with more complex scenarios.
Interpretation of cyclic geometries improved. Now the conceptualization of regular polygons or stars is easier.
Assignment of measures to segments improved. Now the algorithm for assigning drawn measures to the segments is smarter.
Improved the algorithm for resizing the segments without assigned measure when there are other segments with assigned measure.
Fixed a problem in the patch entity when the sides have double points.
Sphere entity generation optimized.
The "Color generator" tool has a new button for clearing the color selection.
The new command "Project to 2D" replaces the selected entities by polynets with the corresponding 2D projection in camera coordinates.
The new command "Solve triangle" calculates the side lengths and angles not provided of a triangle and passes it to the design space.
The new command "Solve tetrahedron" calculates the vertices of a tetrahedron from the set of segment lengths and passes it to the design space.
Minor changes in the sketch interpreter.
Implemented import/export capability to ".stl" format.
The new command "General intersection" calculates the line and surface fragments resulting from the intersection of a set of entities. This new command includes the functionality of the old command "Line intersections" that has been removed.
The new command "Cut-line between surfaces" returns the cut-line resulting from the intersection of a set of surface entities.
The "Browser's storage tool" includes now an "Overwrite" button for each record.
A new command "Generate surface contour" returns the bounding contour of the selected surfaces.
Several improvements in the sketch interpreter.
Some of the initial stages of the sketch interpretation have been modified in order to make the analysis more image-size independent. The previous versions had many threshold values correctly calibrated for small images but not adequate for larger images.
The new command "Copy properties" allows to copy the configured members from the selected source entity to the selected set of target entities.
A new command "Focus entity set" allows to focus the view on the selected set of entities.
The node items shown in the 'Tree explorer' tool includes now a button to focus the view on the rendered entity owning the node.
Several improvements in the sketch interpreter.
OCR function improved.
Imagination of hidden faces improved.
A new command "Line intersections" calculates the line fragments resulting from the intersection of several 3D linear entities.
A set of commands for creating regular polyhedrons have been implemented.
A light OCR (Optical Character Recognition) functionality for recognizing drawn measures has been implemented in the sketch interpreter. Now the user can optionally provide measures to some of the sketched elements in order to obtain a geometry with specific dimensions.
A new function for imagining hidden faces has been implemented in the sketch interpreter.
The following configurations have been added to the sub-menu "Design assistant options":
  • Perform OCR.
  • Imagine hidden faces.
  • Print assigned measures.
  • Print non-assigned measures.
The function for opening drawing files stored in external domains has been enabled.
Several improvements in the sketch interpreter.
A new command "Batch modify" allows to modify, remove or add members to entities in batch mode.
The sketch interpreter implements a new button for defining a constrain window for the sketch analysis.
A new command for creating regular polygons has been implemented.
The new command "Set default properties" allows to change colors and other properties to the selected entities.
The new menu item "Label as 'auxiliary'" allows to create new entities labeled as "auxiliary".
Now the auxiliary constructions are directly created with the label "auxiliary".
The new command "Select auxiliary entities" appends to the current selection all entities with the label "auxiliary". After that the auxiliary entities can be easily hidden or deleted.
The command "Properties" has been renamed to "Physic properties" and now shows at the end the aggregated properties of the selected entities as a compound.
The "mesh" entity type has been improved:
  • The new members "closed s" and "closed t" allow to generate the surface closed in any or both of the two directions.
  • The new member "table colors" is the convenient way to provide the colors for the set of vertices defining the geometry.
  • The new member "smooth" allows to create organic surfaces with an accurate control over the points by which the surface must pass.
A new command for creating mesh entities has been implemented.
The command "Align entities" now includes the option to perform a scaling while aligning the entities.
The "Modify vertex" command allows now to modify also "mesh" and "patch" entities.
The WebGL 1.0 extension "OES_element_index_uint" has been enabled to allow the rendering of geometries with a huge number of vertices.
Fixed problem in the generation algorithm of the "pipe" entity type.
A new command for creating general revolved entities has been implemented.
A new command for creating extruded surfaces has been implemented.
A new command for creating patch entities has been implemented.
A new command for generating polar array distributions has been implemented.
The on-screen point selection function has been improved.
A new set of commands for creating circles and arcs has been implemented.
A new set of commands for creating spheres has been implemented.
A new set of commands for creating cylinders and cones has been implemented.
Now the UI allows the on-screen selection of singular points like centers or control points, apart from standard vertices.
The "Modify vertex" command now allows to modify also "curve" entities.
A new function for entering relative coordinates has been implemented. It allows to define a new point as an offset with respect the previous entered point.
The new command "Convert to wireframe" allows to obtain the wire representation of the selected entity.
A new sub-menu "Edition options" has been created to group those general configurable options applicable to the commands directly involved in the edition and creation of entities.
The auxiliary constructions have been improved with new in-place configurations.
The commands for editing drawing metadata have been improved.
New member "base" added to the "polar" entity type. If provided then the elements are only translated and not rotated as they are copied to the destination points of the polar array distribution.
The "Share" function offers a new configurable option to include the current view definition in the URL start-up instructions. This allows to share the drawing using the current view definition without modifying the own drawing source.
Curve's algorithm optimized.
The "Plan view" command has been replaced by the collection of commands "Predefined views".
The rendering context type indicator button shown in the bottom-right corner has been replaced by the much more useful "Zoom extent" button. The "Zoom extent" menu item of the "View control" menu has been removed.
The menu command "Purge browser's storage" has been moved to a button within the "Browser's storage" tool.
The new command "Optimize drawing" merges all the entities of the drawing into a reduced set of polynets in order to improve the performance of the widget when using WebGL.
The color picker control has been improved.
The command "Merge entities" has a new option that allows to decide if the signals must be preserved in the result or not.
Now the design assistant will show the last used command as a suggestion.
The text responses of the design assistant now have a "copy to clipboard" button.
The command "WebGL properties" has been integrated in the command "Browser capabilities".
Commands for configuring default color for faces and lines in new entities have been implemented.
A new function for pausing a command has been implemented. If used then the current command is paused and the user can change the camera configuration in the usual way. This function is typically used while selecting points in the screen that are hidden for the current camera orientation.
New command "Create text signal" for creating polynets with text signals.
The command "Create polygonal" has been replaced by the command "Create line" for creating 'line' entities.
New command "Create face" for creating 'face' entities.
New command "Create box" for creating 'box' entities.
The "Tree explorer" component is improved by adding a button for inspecting the parent node, and a button for inserting new nodes before the current one.
Now the command "Copy entities" can copy the same selection to multiple destination points.
The "Explode entities" command has been improved including the new in-place configuration "create 'face' and 'line' entities instead of polynets".
A new command "Concatenate lines" has been implemented. It is useful for generating a single line from a group of different connected lines.
The recognition of polyhedrons in the "design assistant" has been improved.
A progressive zoom function has been implemented for the Zoom-in and Zoom-out buttons of the top-right corner. The progressive zoom can be performed applying a long-press event over the buttons. The traditional stepped zoom function is preserved.
Some vector and geometric calculation functions have been added to the JS object "AppLai4d" in order to enrich and facilitate the development of algorithmic entities. See the "program" entity section in the Reference manual for more information.
The command "Merge Entities" allows now to decide whether to merge the vertices or not.
Official file extension for LAI4D drawings is ".L4D".
Implemented import/export capability to ".ply" (point cloud) format.
Implemented import/export capability to ".obj" format.
New member "append" added to the URL start-up instructions. The "append" member is intended to let the user to modify the drawing source from the URL before it is interpreted.
Improved user interface for the creation of polygonal entities.
Icon size of on-screen controls increased.
Initial version.

+         Legal note         Contact         License         Terms and conditions         ↓
The Laboratory of Artificial Intelligence for Design is a not-for-profit initiative aimed to experiment the research results of Lai4d Systems in the field of AI applied to engineering.
LAI4D, Iquix, Gosyx and JAIC are trademarks or registered trademarks of Juan Andrés Hurtado Baeza.
This web site only stores information in the user's web browser for statistical purposes (GDPR).

For administrative inquiries regarding LAI4D, please contact Lai4d Systems.

LAI4D, Iquix, Gosyx, this web site and accompanying documentation constitute a software system (the "Software") subject to the following license agreement. By using this Software, you accept these terms. If you do not accept them, do not use this Software:

The Software is free for use in any environment, including but not necessarily limited to: personal, academic, commercial, government, business, non-profit, and for-profit. "Free" in the preceding sentence means that there is no cost or charge associated with its use. Permission is hereby granted, free of charge, to any person obtaining a copy of the Software, to use the Software without restriction, including the rights to use, copy, publish, and distribute the Software, and to permit persons to whom the Software is furnished to do so.

You shall not sell, rent, lease or sublicense the Software or any part thereof. You shall not modify, adapt, or translate the Software. If you create derivative works based upon the Software or use the Software as a component of other system, provided the Software is not modified, you shall not alter its appearance neither hide its presence or any of its features. If you create a web site offering copies of the Software you shall clearly indicate that it is a mirror site and publish a link to the official web site. You shall not reverse engineer, decompile, deobfuscate, disassemble, or otherwise attempt to discover the source code of the Software. You shall not alter or remove any copyright or other proprietary notice that appears on or in the Software.

The Software incorporates an Internet-based service named COMMUNITY providing a space for cooperation between users and developers. The Software incorporates an Internet-based service providing free online capabilities for file storage and retrieval. The content may be submitted by anonymous users and it is not offered any kind of warranty over it. The submitted content will be publicly available. The integrity or accessibility of the content cannot be indefinitely warranted. The person submitting the content shall be the author or shall have the necessary permissions to submit it. The author of the submitted content renounces to any exclusive right over it that becomes of public domain.

You shall not use the Software, including Internet-based services, for illegal or offensive purposes.

The functions of the Software, including Internet-bases services, may be changed or canceled at any time without previous notice.

If any part of this agreement is found void and unenforceable, it will not affect the validity of the balance of this agreement, which shall remain valid and enforceable according to its terms.

This legal notice shall be included or linked in all copies of the Software.


contador pagina web