Dexter Shepherd

---

My PhD “Get a Grip: Comparing Optical and Electrical Tactile Sensors for Adaptive Robotic Locomotion” at the University of Sussex has focused on constructing various tactile sensors (PCB manufacturing and soft body development) to explore how legged robots can perceive and adapt to the richness of real-world contact. This entailed a number of elements: constructing many tactile sensors including a novel one of my own design, testing them on board self-built robots, constructing equipment to gather large datasets (all openly available); and using a range of ML models of different complexity and temporal aspects such as regression, support vector machines and random forests on the less heavy side, and CNNs and LSTMs on the larger model side to classify this data. As seen in my recent publication “Texture and friction classification: optical TacTip vs. vibrational piezoelectric and accelerometer tactile sensors” I have a strong background in using datasets and performing statistical analysis.

My research has demonstrated that low-cost, low-resolution sensors can still achieve complex perceptual tasks such as texture and orientation detection, challenging the notion that high-resolution sensing is always necessary [1,2]. Alongside sensor development I have worked across simulation environments, and neural control models (including CTRNNs for evolved gait control), bridging hardware and software to create embodied systems capable of adaptive behaviour. Through this work, I have presented at various conferences winning a best paper at UKRAS and best presenter at IEEE AIRC, and contributed open datasets and open-source libraries [3,4,5] that support reproducibility and accessibility in AI research.

My PhD puts me in the rare position of being at the intersection of AI and engineering. I can use CAD, 3D print, laser cut, PCB design, construct electronics and robotic chassis (servos, DC motors, microcontrollers) but also program it. In addition, I have a strong understanding of machine learning/AI concepts including: evolutionary algorithms, neural models and deployment of embedded machine learning on circuits. Outside my PhD I pursue many passion projects (as seen on my Github) looking into new directions and trying to develop new skills in trending AI topics such as large language models and new types of simulators such as MoJuCo.

My knowledge of AI and machine learning is not limited to robotics or tactile sensing. I have coordinated and led teaching projects across Africa, working multiculturally to teach biomedical and agricultural students to construct data logging devices, gather datasets in the field and then use AI/ML to process and analyse this data. Working interdisciplinarily comes naturally to me having worked and collaborated across a university committed to interdisciplinary research.

I bring a commitment to building inclusive and collaborative communities. For several years I led the Peer Assisted Learning scheme at Sussex in which students support their peers, expanding student support and winning consecutive education awards. I also co-founded and co-organise conferences and mentoring programmes that foster exchange between early-career researchers and students across disciplines and continents. This website serves as a living portfolio, a space to document my projects, research, and technical explorations over time.

View CV

Qualifications

Experience

Volunteer Course Instructor / Course Convenor

TReND in Africa, Sussex AI, BioRTC, LUANAR (Malawi), Yobe State University (Nigeria)

Dec 2023 – Present

• Delivered postgraduate machine learning courses in Malawi covering data analysis, regression, classification, and deep learning.
• Led an online Python crash course, coordinating a team of five instructors and developing video content and worksheets.
• Directed African operations in 2025, managing logistics, curriculum design, and deployment of ML and data-logging hardware courses.
• Delivered machine learning workshops for computational neuroscience in Yobe State, Nigeria.

Head of Student Mentor Scheme

University of Sussex – School of Engineering and Informatics

Aug 2022 – Dec 2025

• Led a student mentoring team delivering academic and welfare support sessions.
• Coordinated workshops and provided regular briefings to senior faculty on student needs.
• Doubled student participation during tenure through improved engagement and support structure.

Senior Teaching Assistant

University of Sussex – School of Engineering and Informatics

Oct 2022 – May 2025

• Taught over eight modules in AI, robotics, and computer science.
• Designed assessments involving evolutionary agents for real-world robots.
• Led construction of 200+ Python-controlled robots.
• Developed open-source machine learning and robotics libraries for embedded systems.

Research Assistant

University of Sussex – CoNNeCT Group

May 2021 – Jul 2021, May 2022 – Aug 2022

• Built low-cost 3D-printed robotic platforms using Jetson Nano controllers.
• Investigated ant-inspired navigation algorithms in real-world environments.
• Integrated GPS and mobile signal tracking systems for outdoor robotics experiments.

Reserve Infantry Soldier

British Army – Princess of Wales Royal Regiment

Oct 2019 – Oct 2022

• Completed advanced first aid and combat training.
• Developed resilience, teamwork, and leadership under high-pressure conditions.