Alistair McConnell

Dr. Alistair McConnell


Research Focus & Metrics

Academic Career Length

9Years (2018-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
22 May 2025 in Nature Ecology & Evolution
Verified
Journal Article
Open access
PDF file
SPStephen Pringle
Stephen Pringle
MDMartin Dallimer
Martin Dallimer
MGMark A. Goddard
Mark A. Goddard
LGLéni K. Le Goff
Léni K. Le Goff
EHEmma Hart
Emma Hart
ZDZoe Davies
Zoe Davies
With biodiversity loss escalating globally, a step change is needed in our capacity to accurately monitor species populations across ecosystems. Robotic and autonomous systems (RAS) offer technological solutions that may substantially advance terrestrial biodiversity monitoring, but this potential is yet to be considered systematically. We used a modified Delphi technique to synthesize knowledge from 98 biodiversity experts and 31 RAS experts, who identified the major methodological barriers that currently hinder monitoring, and explored the opportunities and challenges that RAS offer in overcoming these barriers. Biodiversity experts identified four barrier categories: site access, species and individual identification, data handling and storage, and power and network availability. Robotics experts highlighted technologies that could overcome these barriers and identified the developments needed to facilitate RAS-based autonomous biodiversity monitoring. Some existing RAS could be optimized relatively easily to survey species but would require development to be suitable for monitoring of more ‘difficult’ taxa and robust enough to work under uncontrolled conditions within ecosystems. Other nascent technologies (for instance, new sensors and biodegradable robots) need accelerated research. Overall, it was felt that RAS could lead to major progress in monitoring of terrestrial biodiversity by supplementing rather than supplanting existing methods. Transdisciplinarity needs to be fostered between biodiversity and RAS experts so that future ideas and technologies can be codeveloped effectively.

16 October 2018 in Sensors
Verified
Journal Article
Open access
PDF file
MMMohammed Mohammed
Mohammed Mohammed
MNMarkus Nemitz
Markus Nemitz
SASimona Aracri
Simona Aracri
AMAlistair McConnell
Alistair McConnell
RMRoss M. McKenzie
Ross M. McKenzie
Adam A. StokesAdam A. Stokes
Adam A. Stokes
The oil and gas industry faces increasing pressure to remove people from dangerous offshore environments. Robots present a cost-effective and safe method for inspection, repair, and maintenance of topside and marine offshore infrastructure. In this work, we introduce a new multi-sensing platform, the Limpet, which is designed to be low-cost and highly manufacturable, and thus can be deployed in huge collectives for monitoring offshore platforms. The Limpet can be considered an instrument, where in abstract terms, an instrument is a device that transforms a physical variable of interest (measurand) into a form that is suitable for recording (measurement). The Limpet is designed to be part of the ORCA (Offshore Robotics for Certification of Assets) Hub System, which consists of the offshore assets and all the robots (Underwater Autonomous Vehicles, drones, mobile legged robots etc.) interacting with them. The Limpet comprises the sensing aspect of the ORCA Hub System. We integrated the Limpet with Robot Operating System (ROS), which allows it to interact with other robots in the ORCA Hub System. In this work, we demonstrate how the Limpet can be used to achieve real-time condition monitoring for offshore structures, by combining remote sensing with signal-processing techniques. We show an example of this approach for monitoring offshore wind turbines, by designing an experimental setup to mimic a wind turbine using a stepper motor and custom-designed acrylic fan blades. We use the distance sensor, which is a Time-of-Flight sensor, to achieve the monitoring process. We use two different approaches for the condition monitoring process: offline and online classification. We tested the offline classification approach using two different communication techniques: serial and Wi-Fi. We performed the online classification approach using two different communication techniques: LoRa and optical. We train our classifier offline and transfer its parameters to the Limpet for online classification. We simulated and classified four different faults in the operation of wind turbines. We tailored a data processing procedure for the gathered data and trained the Limpet to distinguish among each of the functioning states. The results show successful classification using the online approach, where the processing and analysis of the data is done on-board by the microcontroller. By using online classification, we reduce the information density of our transmissions, which allows us to substitute short-range high-bandwidth communication systems with low-bandwidth long-range communication systems. This work shines light on how robots can perform on-board signal processing and analysis to gain multi-functional sensing capabilities, improve their communication requirements, and monitor the structural health of equipment.

5 March 2026 in International Journal of Social Robotics
Verified
Journal Article
Open access
ABAmani Al Bayrakdar
Amani Al Bayrakdar
Gosha ColquhounGosha Colquhoun
Gosha Colquhoun
MDMauro Dragone
Mauro Dragone
AMAlistair McConnell
Alistair McConnell
RPRuth Paterson
Ruth Paterson
Over 537 million adults live with diabetes, yet many lack sufficient support for effective self-management, resulting in suboptimal glycaemic levels and rising healthcare costs. Socially assistive robots (SARs) may fill this gap through offering real-time, personalised support. This study explored adults’ views on the design, usability, and acceptability of a SAR-based intervention. A three-phase exploratory co-design approach, guided by the Medical Research Council framework, involved 16 adults with type 1 and 2 diabetes in Scotland. Phase 1 involved forming a working group and conducting a scoping review. Phase 2 included two online focus groups to identify challenges, technology use, and desired SAR features; and two in-person workshops at the UK National Robotarium to assess SAR’s usability, acceptability, and integration into daily life. Phase 3 involved thematic analysis, translating user requirements, and planning future feasibility. Participants identified barriers including dietary adherence, insulin adjustments, and limited access to clinicians. Technologies like glucose monitors and insulin pumps were seen as burdensome, highlighting the need for personalised, on-demand support. In conceptualising a SAR, participants recommended interactive coaching features, such as real-time medication guidance, meal planning, emergency alerts, and data-sharing. Priorities included a user-friendly, inclusive design, strong data privacy, and integration with existing devices and care pathways. SARs offer a feasible approach to personalised self-management, with user-driven features like tailored coaching, accessible design, and healthcare connectivity improving glycaemic levels and well-being. These findings inform prototyping and evaluation of a new SAR intervention for diabetes management.

16 October 2026 in Applied Sciences
Verified
Journal Article
Open access
PDF file
SWSiwen Wang
Siwen Wang
AMAlistair McConnell
Alistair McConnell
MVMarta Vallejo
Marta Vallejo
Human visual attention plays an important role in how people perceive and respond to environments containing potential risks. It is unclear whether vision–language models (VLMs) can predict where people look in such scenes. This exploratory study compares human gaze with VLM saliency predictions in safety-relevant scenes. Ten participants freely viewed 33 scene images while their gaze was recorded with eye-tracking glasses. Four VLMs (GPT-4o, Gemini 2.5 Pro, Gemini 2.5 Flash and Claude Sonnet 4) were prompted to predict gaze points for the same images, and the resulting maps were compared with the human gaze maps using standard saliency metrics. All four models showed positive spatial agreement with human gaze on the standard metrics (mean Pearson correlation 0.53 for GPT-4o), but a simple centre baseline scored higher. After correcting for centre bias, three of the four models remained above chance. The results for GPT-4o were stable across repeated runs and did not depend on the instructions in the prompt. Overall, current VLMs can partly predict where people look in safety-relevant scenes from a prompt alone, without any eye-tracking training data, but their predictions remain limited.