Over recent years, we have begun investigating the potential for bringing computer science tools into the design, programming, and deployment of molecular robots, measuing in the nano-meter scale. Nanobots (as we affectionally call them) have been in continuous development all over the world for a number of years. However, essentially all are being developed with a single medical task in mind. In contrast, we want to take advantage of computer science to create new treatments from combinations of existing and new nanorobots. Using compilation, optimization, simulation, information retrieval, and artificial intelligence, the development process can be accelerated immesurably.
Two major challenges we have begun to tackle are (i) automating the safety of nanobots; and (ii) nanobot programming languages:
- G. A. Kaminka, R. Spokoini-Stern, Y. Amir, N. Agmon, and I. Bachelet. Molecular robots obeying Asimov’s three laws of robotics. Artificial Life, 23(3):343–350, 2017.
- I. Wiesel-Kapah, G. A. Kaminka, G. Hachmon, N. Agmon, and I. Bachelet. Rule-based programming of molecular robot swarms for biomedical applications. In Proceedings of the International Joint Conference on Artificial Intelligence, pages 3505–3512, 2016.
While working on the Tolkien molecular robotics project, we realized we were having trouble finding relevant papers: PubMed, google scholar, Arxiv and others were not enough. We therefore built rivendell.cs.biu.ac.il, a new kind of search engine specifically built to find papers, given a context for your project. It uses advanced machine learning and natural language techniques to continually improve the search results specific to the project topics. Check it out!