Evolving embodied intelligence and Use of Robotics in the Management and Conservation of the Pemba Channel Tanzania
Event details
Evolving embodied intelligence
Professor Dr. A.E. Eiben, Vrije Universiteit Amsterdam (NL), University of York (UK)
Evolutionary robotics is the art of employing evolution to develop the brains (controllers), the bodies (morphologies), or both for autonomous robots. In this talk I explain the benefits of evolving real (not simulated) robots for engineering as well as for fundamental research. I argue that constructing systems of self-reproducing machines will lead to a new type of evolution, the Evolution of Things rooted in evolutionary computing, artificial intelligence, robotics, and artificial life with new challenges and opportunities. I will reflect on the Robot Baby Project and the ongoing Autonomous Robot Evolution project (funded by the EPSRC). Finally, I will discuss a long-term research programme with some “grand questions”, possible applications, and future perspectives.
Recommended Reading
- A.E. Eiben, S. Kernbach, and Evert Haasdijk, Embodied artificial evolution: Artificial evolutionary systems in the 21st Century, Evolutionary Intelligence, 5(4):261-272, 2012
- A.E. Eiben and J. Smith, From evolutionary computation to the evolution of things, Nature, 521:476-482, 2015
- A. E. Eiben, Evolving robot software and hardware, In: Proceedings of the IEEE/ACM 15th International Symposium on Software Engineering for Adaptive and Self-Managing Systems, 2020, pp 1-4.
Use of Robotics in the Management and Conservation of the Pemba Channel Tanzania
Kennedy Edeye - PhD Student in the Department of Environment and Geography, University of York
Information on the spatial distribution of habitats and vulnerable species is important for conservation planning. In particular, detailed knowledge on connectivity of marine ecosystems in relation to depth and seafloor characteristics is crucial for any proposed conservation and management actions. Yet, the bulk of the seafloor remains under-sampled, unstudied and unmapped, thereby limiting our understanding of connections between shallow and deep-water communities. Recent studies on mesophotic coral ecosystems (MCEs) have highlighted the Western Indian Ocean as a particularly understudied marine region. Here we utilise an autonomous underwater vehicle (AUV) to collect in-situ temperature, oxygen concentration, bathymetry, acoustic backscatter and photographic data on benthic communities from shallow (<30 m) and mesophotic (30–150 m) depths at selected sites in the Greater Pemba Channel, Tanzania. Further, we use generalised additive models (GAMs) to determine useful predictors of substratum (hard and sand) and benthic community type (coral, turf algae, fleshy algae, fish). Our results revealed the presence of a complex seafloor characterised by pockmarks, steep slopes, submarine walls, and large boulders. Photographs confirmed the presence of MCE composed of corals, algae and fishes on the eastern margins of the Pemba Channel. The GAMs on the presence and absence of benthic community explained 35%–91% of the deviance in fish and fleshy algae assemblages, respectively. Key predictors of the distribution of hard substrata and the coral reef communities were depth, showing the upper boundary of MCEs present at 30–40 m, and seafloor slope that showed more occurrences on steep slopes. The upper 100 m of water column had stable temperatures (25–26 ◦C) and oxygen concentrations (220–235 μmol/l). We noted the presence of submarine walls, steeply inclined bedrock, which appeared to support a highly bio-diverse community that may be worthy of particular conservation measures. Our results also highlight the capability of using marine robotics, particularly autonomous vehicles, to fill the knowledge gap for areas not readily accessible by divers or with surface vessels, and their potential application for the initial survey and subsequent monitoring of Marine Protected Areas.