Search Engine Working: How
I wrote my Master’s thesis on free speech, the social dimension of capitalism, and unequal lives via robotics. I’ve been studying robotics for over five years, from the beginnings of a Technical University of Munich graduate student degree to my role at the German Federal University as the personal assistant to the Head of the Robotics Hall of Fame, and at the European Robotics Institute at Hamburg University of Applied Sciences. The aim of my recent research is based on the philosophy of “robust robotics”, where the design of the device is fundamentally a methodological design. To this end, I hope my interview series will be useful for others, who are interested in robotics and/or investigating functional design as the main area of study (vocational) in machine learning, analysis (machine vision and artificial intelligence), and computational learning.
In your opinion, what makes for a transformative engineer? Is it the ability to combine machine power with human creativity or more so, the business that it brings?
While I’m often asked these questions, it’s a little arbitrary and less about creativity than technology. I believe that, in a research setting, these processes can flourish when they work together.
The current state of it is, for me, rather depressing as described by Elon Musk. There are very few prototypes and very few projects in production, for a technical reason (cost). Also, there is a clash of cultures between technical researchers and venture capitalists. With great enthusiasm about a design challenge (i.e. in our line of work, an answer to a questionnaire or a prototype), software programmers and researchers have little time to meet their goals as they are forced to eat and sleep the data they produce.
Do you believe the time has come to boost access to basic technologies and tools? What barriers are preventing this kind of shift from a technical standpoint?
The issue of basic materials (read: the capital stock necessary for prototyping and producing commercial products) is rather new. Up until very recently, making a machine showed up on a technical research to-do list, and hence, not on a corporate research agenda. Indeed, in the EU robot research and development (RRD) portfolio, six out of twenty projects and activities fall under the description “basic materials”, as outlined by the European Investment Bank.
The basic materials already exist and are widespread worldwide, however, industrial standards (i.e. standard machine-building materials) are either very complex and need a lot of time and funding to develop a new market for them or very simple (i.e. the building blocks are commercially available). In practice, minimal requirements for basic materials exist from user requirements but can be quite hard to apply effectively due to a lack of standards.
In looking at basic materials as a technical challenge, I think fundamentally the correct approach is to identify the industrial needs and develop such criteria first — a “people’s definition” and an “industrial target” — before coming up with commercially-viable production parameters. This way, even a potential simple product can be commercialized by the adoption of the small number of “advanced” specs that will inevitably prove fundamental for the product — this is one factor that differentiates a commercial success from a technical problem. Furthermore, a high degree of automativeness cannot be done with access to basic materials, so much focus should be placed on gaining access to alternative, specialized materials (preferably from regions with small strategic or energy resources, such as Romania or Hungary, which are currently under consideration for using highly recyclable plastic for foundations and testing packaging).
As a researcher, what do you consider a most successful (technology/product) process?
AI/machine learning is unquestionably something for the future, but it is also possible to make great progress within the horizon — I would like to highlight the success of Starfox, which I was involved in the launch. Indeed, when the project team wanted to build a microphone that could “talk” to people, they had no idea whether they could have a working microphone — they ended up for a prototype — and only later found a mic that actually worked.
The main reason for this “success”, I think, is twofold: there was a very hands-on team that worked on specifications on a regional scale and a collaborative, multidisciplinary effort in terms of software. Moreover, the approach greatly underlined the important differences between the project team and marketing forces. In a nutshell, I can say I think that the approach successfully combined a “small teams approach” with a “multidisciplinary approach” in the manufacturing phase of the project.
More in you, Val Badulla, Contributor at Arabic Research Institute at MSU. Please share your articles, research, and learn how you can bring a global perspective to your research as a research platform.
