Editor Insight: Interview with Professor Marcelo H. Ang, Jr. 

Artificial intelligence; machine learning specifically. This is an emerging trend: how to make robotics more useful is learning from data, learning from experience. There’s a lot of work being done right now in that area, basically moving away from traditional scientific based or model-based algorithms to make robots more capable. The first use of that that we can see now is self-driving vehicles. 

Research focusing on robotics in unstructured and human environments. How do we get robots to do this, and how do we learn from experience? Can robots learn from experience, like how humans learn from experience? There are topics such as learning from demonstration or observing human behaviour and trying to abstract the skills needed to do that.

So, I’d like to encourage more of that. Robotics today can move very well, control forces very well. But is that enough to open a water bottle, for example? There is a missing gap between high level tasks and lower-level motion and control capabilities of robots. This is the holy grail of robotics – how to bridge the high-level tasks into motion and force action primitives: aka the lowest level.   

As an author, when you plan to publish something, first ask yourself, why do you want to publish it? It’s likely you want to publish something new. The biggest challenge is identifying what is that ‘new’? It can be a new theorem, a new theory, a new way of doing things or a better performance of something. It is then important to have in mind what are the contributions of the paper. The problem of many rejections is that this aspect is not made clear. Many of these papers just say ‘I’m solving this problem’, or ‘I did this’ and that’s it, that’s the solution. But how does that compare to the state-of-the-art? 

To do that, it must have a very clear problem statement. That’s also the second most important aspect– after the contributions, the paper must address a clear problem statement – what is the paper trying to do, to share what information; to solve what problems? And why is that still a problem in spite all the publications available. If that is clearly articulated, it will lay the outline of the of the paper, demonstrating that the contributions are real and significant.

Have a clear title and abstract, as these will be the first things we will read. Additionally, ensuring that your submission is easy to read, put yourself in the readers’ shoes. Have a clear contribution, stating what is the novelty of your paper. Sometimes, it doesn’t hurt to get your paper read by someone first (before submitting), to make sure your message is clear.