Wednesday, 19 October 2016

PRiME workshop on PostHumanism and Resilience thinking robots @OUUK @robotics

Today I have the pleasure of attending the Posthuman Resilience in Major Emergencies (PRiME) networking event organised by the OU, UK. This is definitely a timely event as it launches a constructive idea exchange with regard to what we need to think about to enable societies to be resilient in case of major emergencies (natural and human disasters affecting small to big regions). The main aim of the workshop is to bring together researchers and stakeholders from a variety of fields within the future technologies area.

The workshop focuses on emergency situations, particularly in major events and disasters, which in today’s connected world require sophisticated responses involving extraordinarily close collaboration between humans and technologies. The concept of resilience has been identified as encapsulating a highly desirable characteristic of both humans and technologies in these settings. Although resilience has been the subject of extensive research in various academic and technical domains, it needs to be thoroughly re-examined in relation to the prospect of a post-human future, e.g. in 50 to 100 years, in which human capacities may be manipulated and radically enhanced. If you are interested in this challenge and have relevant ideas or expertise, you are invited to join us in our upcoming workshop where the concept of resilience will be a core aspect.

A posthuman approach to resilience might analyse networks of which humans are only a part, or assemblages composed entirely of non-humans. It may involve applying abstract concepts of resilience to humans and nonhumans alike; or "pluralizing" the concept to acknowledge different ways in which things or subjects can exhibit resilience. It may explore the contribution of nonhuman actors to forms of stability traditionally viewed in human terms, or seek greater recognition of diverse interests in being resilient.

The day is filled mostly by 30 min keynotes on posthumanism, resilience, human-machine interaction, communication, and robot technology.

Some first thoughts picked up while liveblogging:
Resilience some info (came in a bit after start of first keynote, train travel).
From Mars exploration, space technologies, self-riding rovers and cars. No external location info.
Use AL mapping area and computer vision & cameras. Mapping the world in 3D, mapping where the rover is, and than plan.

Energy is limited: solar power in combination with battery. Autonomous sensors will use battery power, the more watt’s used, the less energy for moving around. Sometimes cheap sensors can be used, but sometimes (e.g. challenges met) more expensive sensors need to be used. So what I tried is modeling the terrain and looking at which type of sensors can be used. Where the software is going to calculate which sensors can be used in terms of energy investment. Anyway mapping the way as it is explored. In a GPS void environment some mapping and exploration can be one, with additional energy saved. But mapping has it limits as the exact photograph taken will provide detailed information, but as soon as the video angle is different, different information will be given. So, how can different pictures ensure accurate information, build from different sensors. The mars technology is now used in tunnels, surveying tunnels and mapping them. VR, AR tech coming from the 3D models sent out from the tunnels, decreasing the risk for humans. But a major challenge is the data coming out of these 3D models. Too much information to calculate. Deep learning is an option, fueled by theoretical information, and lots of gaming industry feedback. Steep and rapid change, every 6 months giant leaps forward. Using AI to augment, improve and replace human actors. Current state of the art is changing so rapidly, that it exceeds information coming out (papers, tech…). Up to 2010 error rates were high, with deep learning, the errors have come down, and very complex images the machines are classifying better than human beings are doing; this can be used for any visual analysis at the moment and be used to looking for information of interest.

Autonomous robotic for surveillance, that way minimize risk for humans and visualise or provide detailed information, plus dealing with problem of lots of data. 
Another big problem is the human-machine interaction, as the technology (now) does not understand the human communication. The interface to communicate with human/machine. 

(inge: makes me think off a lot of internet of things problems revolving on energy versus tech action. )