Visualizing strategy

Academic 2008–09. In partnership with Komatsu Forest AB (Valmet), IDEO, SLU, and Oryx.

Laser-based system for wood-logging harvesters

Laser-based system for wood-logging harvesters

The ‘Komatsu project’, as my classmates and I used to refer to it, was our first 10-week project at Umeå Institute of Design… And we had to deal with a big machine: the wood-logging harvester.

Of the 10-week duration of this project, the first five were spent doing group research, and the last five we spent developing our individual concepts.

harvester and benjamin

A Valmet 941.1 wood-logging harvester with Benjamín standing beside it’s 2 ton cutting head.

The subject of this project was ‘Complex interactions in specialized vehicles’. Quite fittingly, studying (interaction) design in Sweden, it’s normal (almost de facto) that the main object of our study was the harvester operators, not the machines.

We were fortunate enough to be able to visit actual harvester operators in the forest twice, visit a school that teaches young people how to become operators, try a simulator (Oryx), visit the Umeå Valmet factory, and had a most interesting lecture about the past, present, and future of wood-logging in Sweden (SLU). Three of us even got to present our project at the SIDeR 2009 conference in Eindhoven, The Netherlands.

Research

Cutting trees.

Entering data in the computer.

We, as a class (comprised of 12 people at the time), decided to work as one big unit for the five weeks of research comprised within this 10-week project.

We then divided the research into four different—but, quite naturally, closely related—themes. These themes were: the working environment, perception (semantics), ergonomics (anthropometrics), and decision-making. The sub-team I was part of was focusing on the working environment.

Dennis the operator

Dennis the operator. He is Valmet’s test driver. Basically, he tests all that goes in such a big machine: from new outer parts to a new graphical user-interface to manage the Maxi™ software on the computer inside the cabin of these machines.

My personal research

Already, during our visits in the forest and through our ongoing research, it quickly became clear to me that I would like to work on visual cognition. I read numerous papers on the subject, but the two that inspired me the most were:

Wästlund, E. (2007). Experimental Studies of Human-Computer Interaction: Working memory and mental workload in complex cognition, Department of Psychology, Gothenburg University, Sweden.

Rosch, E. (1978). Principles of categorization in Cognition and categorization, Erlbaum, Hillsdale, NJ.

Although the book by Eleanor Rosch is quite the famous classic, it reminded me of basic design exercises I had done during my BA in industrial design. Even though these ‘gestalt’ are quite general if not generic, they formed what would be the first parcel of idea that would evolve in the braoder Visualizing strategy concept.

Below are quick (and dirty) sketches I made to flesh out my ideas.

sketch

Tremendously rapid—and lousy—sketch of a way to obtain a ‘top-view’ of (a part of) the forest: an inflatable blimp-like balloon with a video-camera attached to it would drift in the skies and give a much more accurate (then say, Google Maps) visual representation of the forested area that will be harvested.

gestalt 1

Ideal situation.

gestalt 2

A more ‘realistic’ representation.

The idea with the above sketches was to simply find an alternative way of representing visual information to the operators. The whole idea with this was to enable them to build meta strategies (where to go next in the forest) more easily.

The preliminary outcomes of my personal research

The ultimate goal of this 10 week project was to increase the productivity of forest harvester operators—and a personal objective of mine, even if utterly hypothetical, was to do so while making these operators jobs more enjoyable as well.

Two premises

1) By reducing the mental workload of the operators. This is where the aforementioned organization—or in other words, order and classification executed by the human brain—kicks in. The idea of providing either tools or ways for the operators to either create order for themselves both in their thought process and reduce their mental workload.

2) If the operators have to focus less on the technicalities of doing a lot of their work “manually”—or, in other words, if their work is truly more automated—, the focus could shift more towards building strategy—the visualization of strategy. One way would be to simplify the management of visual input sources they receive information from… and by directing the focus of their work (cutting trees and piling them) directly where it belongs: in the forest.

Early concepts

Below are two (again) very fast drawings that are at the genesis of my Visualizing strategy concept.

A very early sketch of the concept.

Lasers being projected from the front of the vehicle.

Lasers being displayed on a tree.

Concept refinement

Low-tech approach: the laser.

The “technology factor” isn’t as big a concern for people that are as acquainted with the latest technologies. By implication, if the operators feel less controlled or ‘dominated’ by technology, it would perhaps enable a freer flow of creativity that would hopefully transcend into their work and the strategies that they might employ.

By using a laser-driven technology, we obtain numerous benefits like:

True augmented reality;

Tried, tested and true technology;

Robust;

Cheap but effective;

Precise;

Good visual contrast in virtually any weather conditions.

Reducing by augmenting

Even if the above statement seems contradictory, it made—and still makes—perfect sense to me. Indeed, I figured that with a simple, low-tech approach that uses lasers as a means of displaying information, I was if fact truly—even if crudely—augmenting reality. I was also reducing the quantity of displayed information to the essential: the information regarding the trees specie, quality as well as a number of dots that would indicate, for example, if there would be five of these dots, five different would be displayed directly on them, hence my concept of interface-in-action.

Discreet choice

Another interesting thing to mention with having five ‘dots’ and a line displayed on the tree trunk is that, contrary to the actual system which consists in the operator having to move the cutting head of his harvester approximatively to the desired location (with a pressure-sensitve knob), the operator is now discreetly indicating the exact location where the saw will cut the tree. This approximation is completely absent with my approach, for the operators need only to be pressing a single button (x number of times) depending to which ‘dot’ they want to move the cutting head to, hence providing a discreet input to their machine.

Furthermore, all the information needed to cut the logs to length is, as previously mentioned, displayed directly on the tree. In other words, this means that the operator, once he has set his personal preferences (e.g. the spacing between the dots that are displayed on the tree, the number of dots, etc.) would been able to constantly keep his eyes in the forest.

Prototyping

prototype studio

The final version of my working prototype. Big thanks to Camille Moussette for his help.

During the developmental phase of the project, I felt the need to ‘tangibilize’ it’s actual laser aspect. The usage of lasers seemed quite nice as a concept, but I felt like I had to actually see the lasers being projected onto a surface, and ideally, simulate the movement in some way.

Here’s a video and a few images showing the second iteration of my prototype. It was a succesfull attempt at using sound as an actuator.

1st iteration: projecting a small laser-beam on a mirror mounted onto a step-motor.

2nd iteration: mounting a mirror onto a speaker head.

2nd iteration: deflecting the laser beam on the vibrating mirror.

2nd iteration: creating a ‘laser line’ through sound-based actuation.

Building a more elaborate version of the prototype.

At this point, the next step was to build something slightly more solid, but more importantly, to find a way to simulate the movement of the ‘laser line’ on a tree log.

Although this prototype in itself wasn’t by any means revolutionary, it was a great learning experience for two very important reasons: it was the first time that I sketched in hardware—and learned some more about Arduino and repurposing old electronics—, but also because it helped me—and those I’ve showed my project
to—visualize my concept much more effectively.

Final concept

At this point, I felt like the working sketch/prototype that I had made solidified my concept. I could then go forward in making a video simulation of what it would be like to work with the Visualizing strategy system for an operator.

Below is a video demonstration of the final concept. Sorry if it is a little jittery—it was hard to even stand up behind Dennis in the cabin!.

I would like to sincerely thank our tutors for this project—namely Dario Buzzini, Mattias Andersson, and Niklas Andersson—for their great insights and support.


Report

For those who are interested in reading some more about the project, contact me and I will gladly send you a copy of it.

SIDeR conference.

Back in early February of 2009, I submitted a paper to the SIDeR 2009 “Flirting with the future”… and it got accepted!

The paper was largely based on the Visualizing strategy project. You can read it here [PDF 180 kb].

A proof that I was there!