Like many in the technology industry, I am a big fan of science fiction films and I’ve written in the past about how exciting it is for me to have a job where science fiction and reality literally meet. Over the past few months, several key projects from various Maple and MapleSim users caught my attention for various reasons and once again, I was forced to giggle publicly as the shear cool factor of these applications overcame my normal mature demeanor.
As part of a recent return visit to Japan, I once again had the pleasure of visiting the famed Takanishi Laboratory of the prestigious Waseda University in Tokyo (Director, Dr. Atsuo Takanishi). The Takanishi Lab is world-renowned for its work in humanoid robotics and more recently, with medical applications of mechatronic systems. The Takianishi group has been an active user of Maplesoft products for its model development for many years, initially with Maple but more recently, exploring new possibilities for modeling advanced dynamics for robotic systems. See the Lab’s homepage for some amazing photos and description of their work. Their robots, of course can walk (seems so old-school these days), but they can also play musical instruments, carry human beings up a flight of stairs, and even express emotion in their facial features.
Although humanoids are often considered by Western engineering communities as somewhat eccentric and not entirely practical applications, in countries like Japan and Korea, research into humanoids is a highly competitive and prestigious endeavour. For example, many Japanese auto firms maintain an active and well-funded humanoid research group. We have all seen Honda’s Asimo. As it turns out, Toyota is also very active and has ongoing research collaborations with the Takanishi Lab.
Although government and university officials tend to point to future personal service robots as a strong motivator for this research, there seems to be an even bigger picture fuelling this activity. Dr. Paul Oh, Director of the Drexel Autonomous Systems Laboratory (DASL) comments that the humanoid research and developments constitutes a modern “space race” for Japan, Korea, and other nations. It is a highly visible articulation of technological capacity and creativity. Dr. Oh also points out the potential of such futuristic work to motivate young generations is key to this work. This is likely very true for a generation that grew up with C3PO and R2D2.
It’s been well-publicized that Maple had made a significant contribution to the development of the original dynamic model equations for the current generation of space manipulators used in the International Space Station and the Shuttle program. With the introduction of MapleSim, I’ve been thrilled to see those techniques evolving. Indeed, the descendents of those first brave souls trekking to modeling places that no man or woman had gone before, are successfully deploying the richer MapleSim modeling platform for the next generation space applications.
One of the most prominent among these new applications is a power management simulation platform currently being developed by the team lead by Dr. Amir Khajepour of the University of Waterloo. Dr. Khajepour holds the Canada Research Chair in Mechatronic Vehicle Systems. Sponsored by the Canadian Space Agency, Dr. Khajepour’s team is developing a terrestrial high-fidelity simulation platform for optimizing the power performance to the next generation extraterrestrial robotic rover vehicles. These vehicles are slated for missions to Mars or the Moon over the next decades. The envisioned simulation platform will allow mission engineers to determine the ideal path for the solar powered electric vehicles to travel to a mission destination using the minimum power. In the extreme environments of other worlds, power consumption is everything. Taking the wrong path and exerting too much effort to climb a hill and running out of battery power can result in irrecoverable mission failure.
Dr. Khajepour’s team’s first results of MapleSim-based models for rover dynamics
In many ways, Dr. Khajepour’s team’s MapleSim application stands as one of the most complex to date. The range of modeled subsystems is extensive ranging from rover dynamics, to tire/contact/terrain, motors and peripherals, solar electric components, environmental influences and more. Additionally, the team is using MapleSim’s unique optimized code generation tools for various HIL experiments. Overall, the models are integrated into an innovative genetic optimization scheme that will allow mission engineers to define scenarios and quickly determine optimal mission configuration for the many parameters.
As interesting as all this is, for me, the most exciting thing is not so much the results so far but the implications on the methodologies of complex system simulation. In some general sense the methodology that the team is deploying is not fundamentally different from the techniques used by the pioneers for the first generation of space manipulators – use symbolic computation to derive dynamic model equations, and then use code generation for fast simulation. The big difference is that the Khajepour team used pretty much off the shelf software – i.e. MapleSim, Maple, plus some available add-ons and was able to literally “drag-and-drop” their way to realistic models and optimized code. In contrast, the first generation teams did a fair bit of Maple programming to the extent of creating their own tool chain called SYMOFROS to standardize the methodology. In the end, the core time-saving benefits of the symbolic approach still saved them months of time, there was still a fair bit of manual intellectual labor that went into those early projects. Fortunately, their sacrifice and insight heavily influenced the design and development of MapleSim and today’s generation of space engineers are able to see much further standing on their predecessors’ shoulders. Sorry, I tend to get melodramatic on occasion.
It has been a remarkable couple of decades for robotics. The romantic image of intelligent, useful, powerful, friendly, and occasionally evil, robots was a main theme in our sci-fi culture from the mid twentieth century onward. Then we saw the emergence of flexible automation robots for manufacturing applications take hold in the 1980s for for a while, the concept of robots was, for many of us, somewhat boring and overly practical. As computing technology such as Maple and MapleSim blew open the doors to new possibilities in so many fields, roboticists are now enjoying a new prominence with new initiatives that are truly amazing. The smarter modeling technology continues to be a key part of this revolution as the triple threat of decreased model development time, increased model fidelity, and faster HIL performance seems to be so much more important to the robotics field than other engineering communities.