Bot & Dolly: “Box” Interview and Behind the Scenes

As a follow up to our earlier post regarding the debut of Bot & Dolly’s mind-boggling short, “Box,” (above) we’re sharing an interview with the team as well as a behind the scenes video produced with The Creators Project.

Behind the Scenes

This behind the scenes film contains interviews with GMunk (Bradley G Munkowitz), BOX’s Design Director and Tarik Abdel-Gawad, BOX’s Creative Director together with behind scenes footage from the making of BOX.

Interview with Bot & Dolly’s Tarik Abdel Gawad, Creative and Technical director on “Box”

Can you please confirm for our readers that none of the box visuals were comped in post?

Yes, this is a capture of a physical performance. The visuals are not added in post.

Where did the idea for this project come from? Was it commissioned?

Box is an internal project that grew naturally out the the intersection of art and technology at Bot & Dolly. We have a great interdisciplinary team of designers and engineers that made the project possible.

From the start, the exploration of classical magic fit with our creative process. Magicians have a long history of mixing technology with performance and the categories of classical magic were perfect inspiration for the geometric illusions in Box.


Can you tell us a little more about the robots? What are those robots normally used for?

The spec sheets on the Bot & Dolly website are the best source of information on our robots.

IRIS Spec Sheet
SCOUT Spec Sheet

How did you work out the choreography between the performer and the robots?

Working out the choreography was a process of rehearsal and iteration. For mainly practical reasons it was actually me performing. I had the most experience operating the robots, and since this was an internal project, rehearsals often took place at night. Each robot weighs around one and a half tons so it takes awhile to get comfortable moving around them, and safety is important.

How did you track the movement of the surfaces by the projector? Was it all preprogrammed based on the robots’ movements?

The projectors and robots are all calibrated within the same coordinate frame. Bot & Dolly’s software, BDMove, makes its possible to synchronize graphic content with robotic motion.

What was the design process for this like? Where did you start? And did you need to test and iterate a bit before getting it down?

We would start with a category of classical magic and begin exploring limited narratives made up of only abstract geometric shapes and a single performer. During the animation phase we relied on quick hand drawings and moving blocks around a table to communicate ideas because it was nearly impossible to describe something just with words.

Choosing the right geometry was very important to creating the illusion of depth, and directly affects the robotic motion. The primary illusion is created by transforming the geometry of the physical 4’ by 8’ canvas mounted to the robotic arm, through projection.

In the first section, “Transformation”, we extrude the canvas into a cube. Later on we combine two canvases to form a larger hinging shape, which in return affects the robotic animation. We tried to make each section build upon the last, and we were always learning something based on what was just completed. By the end, we ended up with a very complex environment, the performer is inside of the projected volume, there are holes in the floor and line drawing on the back wall.

What was the most challenging aspect of the job?

The difficult part is that you don’t know exactly how something will turn out until you’ve seen it projected in the space. Even then it changes with the environment’s lighting, which is also synchronized with the graphic content and robotic movement.

Many software applications were used on this project in conjunction with BDMove. It’s a very collaborative process with a lot of creative control, so it takes a lot of time and iteration to get to the point where everyone is happy.

What’s next for you?
I’m not sure. One of the things I love about working at Bot & Dolly is that we tackle a wide range of problems, both on the creative and technical side. There are applications in a variety of fields for the technology demonstrated in box, which makes it hard to predict the next thing.

Tags: , , , , ,

About the author

Justin Cone

Together with Carlos El Asmar, Justin co-founded Motionographer, F5 and The Motion Awards. He currently lives in Austin, Texas with is wife, son and fluffball of a dog. Before taking on Motionographer full-time, Justin worked in various capacities at Psyop, NBC-Universal, Apple, Adobe and SCAD.


Met Hrovat

I was so hoping to see some behind the scenes. Fantastic, thanks!

Marcus Andrew

Amazing stuff.

Peter, Reflex Motion Control, South Africa

A very interesting use of not 1 but 3 motion control rigs! And integration of the 3 robots with the projection software…industrial robots have proprietary software that does not mix easily with the likes of Maya, etc. The main issue is how cameras, or in robot’s language,tools, are dealt with in 3d space. Where the camera nodal point, and in this case the center point of the 4×8 is the easier part, but getting the angle of orientation right is the tricky bit. Robots use rotational angles, whereas graphics platforms tend to use target locations for where the camera is looking. Defining this and writing translational scripts to hande this is where these guys ave really pushed the envelope!

We did a similar thing with Houdini and a robot to shoot 3d background plates with a single camera and then integrate with CG created in Houdini. Two commercials from 3 weeks of extended days in studio…

I am pleased at last the film industry is taking industrial robots seriously. The bridge between these beasts and the CG tools you all love to play with is being built, and by using robust robotics the reliability is improving our motion control industry’s reputation which has suffered due to the use of unreliable home grown systems.


“Yes, this is a capture of a physical performance. The visuals are not added in post.”

I saw very little of the actual live projection in the making-of video. Only a millisecond which produced a quite dirty effect (light bleeding onto the walls). I don’t quite see how the projections are being projected onto the canvases – but I will take their word for it that they are – but I do not believe that this film was not comped together – even if they take a clean pass of the move and use it to create crisp edges. If that is the case, and forgive me if I am wrong, then to say the visuals are not added in post seems disingenuous.


Out of this world! .. speechless


Great article, thanks for sharing! Added a link of this on my website:

Comments are closed.