Publications home
Analysis of ‘quantified-self technologies’: an explanation of failure
at
| Cécile Boulard, Tommaso Colombino, Antonietta Grasso |
| 20th Congress of the International Ergonomics Association (IEA), Florence, Italy, 26-30 August, 2018 |
| Download |
@inproceedings{boulard2018analysis,
title={Analysis of ‘Quantified-Self Technologies’: An Explanation of Failure},
author={Boulard-Masson, C{\'e}cile and Colombino, Tommaso and Grasso, Antonietta},
booktitle={Congress of the International Ergonomics Association},
pages={579--583},
year={2018},
organization={Springer}
}One of the driving principles of the quantified-self (QS) movement is that knowledge is power. To have fine-grained and objective measurements of our body and its functions, and of our routine activities should, in theory, give us better control over them. But what happens when quantification gives us a representation of ourselves that we don’t understand? Do we question the quantified model of self, its objectivity and accuracy, or do we question ourselves? Drawing on the two ergonomic studies of QS inspired technologies, we want to provide some reflection on why there may be a mismatch or misunderstanding between measurement and self-representation.
The technologies considered in this paper are an activity tracker counting the number of steps in a day and a solution for daily commuting estimating the CO2 emissions and costs associated to commuting practices. The methodologies put in place in this studies are interviews, diaries and figures gathered by the trackers. The results depict that the technologies and the information they provide are not fully accepted by the users. The main raison seems to relate to a mismatch between the reductionist way the technologies present potentially complex issues and the users’ understanding and self-perception in isolation.
The activity tracker is essentially a gamified pedometer, which reduces the notion of fitness to a step-count, and the notion of improvement in fitness to the attainment of arbitrary, incremental goals. The use of such trackers may be useful within the context of a health intervention targeted to a user. On its own, however, the activity tracker tends to give users the perception that it provides an “unfair” characterization of their efforts and progress.
As regard commuting practices, the goal was to motivate users toward greener practices by providing metrics on CO2 emissions and costs of alternative modes of transport. One of the problems with it was that the individual CO2 footprint is not only difficult to calculate with accuracy, but may also be counterintuitive when provided in a comparative way across transportation means and provide unexpected feedback to people that are consciously making an effort to reduce their environmental impact. This might actually discourage them from making an effort to reduce CO2 emissions.
Overall we found that with QS technologies, there is a risk of decontextualizing and reducing complex activities to simple calculations which encourages binary true-false thinking on the part of users. This leaves little room for a nuanced understanding of the underlying problem and of the specific circumstances and requirements of any individual user. We would like to propose that quantified-self technologies may benefit from less simplified models, even at the expense of more complexity, but be able to provide more contextual and ultimately understandable quantifications for the users.
To make robots autonomous in real-world everyday spaces, they should be able to learn from their interactions within these spaces, how to best execute tasks specified by non-expert users in a safe and reliable way. To do so requires sequential decision-making skills that combine machine learning, adaptive planning and control in uncertain environments as well as solving hard combinatorial optimisation problems. Our research combines expertise in reinforcement learning, computer vision, robotic control, sim2real transfer, large multimodal foundation models and neural combinatorial optimisation to build AI-based architectures and algorithms to improve robot autonomy and robustness when completing everyday complex tasks in constantly changing environments.
VISION
The research we conduct on expressive visual representations is applicable to visual search, object detection, image classification and the automatic extraction of 3D human poses and shapes that can be used for human behavior understanding and prediction, human-robot interaction or even avatar animation. We also extract 3D information from images that can be used for intelligent robot navigation, augmented reality and the 3D reconstruction of objects, buildings or even entire cities.
Our work covers the spectrum from unsupervised to supervised approaches, and from very deep architectures to very compact ones. We’re excited about the promise of big data to bring big performance gains to our algorithms but also passionate about the challenge of working in data-scarce and low-power scenarios.
Furthermore, we believe that a modern computer vision system needs to be able to continuously adapt itself to its environment and to improve itself via lifelong learning. Our driving goal is to use our research to deliver embodied intelligence to our users in robotics, autonomous driving, via phone cameras and any other visual means to reach people wherever they may be.
This web site uses cookies for the site search, to display videos and for aggregate site analytics.
Learn more about these cookies in our privacy notice.
Cookie settings
You may choose which kind of cookies you allow when visiting this website. Click on "Save cookie settings" to apply your choice.
This content is currently blocked. To view the content please either 'Accept social media cookies' or 'Accept all cookies'.
For more information on cookies see our privacy notice.