Periodic Reporting for period 3 - PROBOSCIS (Proboscidean sensitive soft robot for versatile gripping)
Reporting period: 2022-08-01 to 2025-04-30
The elephant’s proboscis serves as an extraordinary biological model, inspiring a groundbreaking concept in robotic manipulation. Elephants are capable of performing both powerful and delicate tasks with exceptional precision, thanks to their trunk’s unique structure—comprised of over a hundred thousand muscles and protected by durable skin.
The core idea behind PROBOSCIS was to draw inspiration from the anatomy, morphology, and gripping behavior of the natural trunk to develop innovative soft robotic strategies for versatile manipulation. This approach aimed to establish a new paradigm in tactile-based, universal robotic manipulation.
The project culminated in the creation of a fully integrated, bio-inspired continuum robotic manipulator. This prototype eliminates the conventional separation between arm and gripper, achieving the versatility of an elephant’s trunk through whole-arm grasping.
The core idea behind PROBOSCIS was to draw inspiration from the anatomy, morphology, and gripping behavior of the natural trunk to develop innovative soft robotic strategies for versatile manipulation. This approach aimed to establish a new paradigm in tactile-based, universal robotic manipulation.
The project culminated in the creation of a fully integrated, bio-inspired continuum robotic manipulator. This prototype eliminates the conventional separation between arm and gripper, achieving the versatility of an elephant’s trunk through whole-arm grasping.
The PROBOSCIS project achieved a breakthrough in continuum robotic manipulation through an innovative, bioinspired approach. Key technological advancements included the development of novel biomimetic artificial muscles, distributed soft and durable tactile sensors, and flexible yet robust materials. These were complemented by new fabrication processes tailored for large, compliant components. The project’s integrative design—eliminating the conventional separation between arm and gripper—enabled the construction of a continuum robotic manipulator that replicates the versatility of an elephant's trunk through whole-arm grasping.
Bioinspired control strategies were also explored, drawing on stereotypical elephant movements. These strategies leverage environmental interaction and tactile feedback to provide the robot with easily learnable models, bodily awareness, and internal representations of its surroundings. This work supports the development of intuitive, adaptive robotic behavior.
The project contributed to additive manufacturing through the creation of new functional material formulations and advanced production techniques, resulting in commercialized products.
The artificial trunk’s versatile grasping capabilities pave the way for novel applications in both academic and industrial research focused on dexterous robotic manipulation. In the long term, PROBOSCIS has the potential to evolve into a ‘living’ assistant, easing human-robot interaction in challenging contexts. It could assist with tasks such as aiding mobility, lifting loads, or interacting gently with patients—especially valuable in healthcare and assistive environments. In search and rescue operations (e.g. after earthquakes or hurricanes), the robot could carefully remove debris and identify victims even in low-visibility conditions like darkness or smoke, with sufficient strength to safely move people to secure locations.
Industrial applications are equally promising. In the food industry, PROBOSCIS could handle diverse items and assess their condition without relying on vision, particularly when objects are occluded during manipulation. It also offers solutions for complex tasks like picking and sorting objects of various shapes and materials—highly demanded in e-commerce, warehousing, and logistics—or manipulating deformable materials such as paper or fabric.
The project resulted in the publication of: 27 papers in open-access (OA) journals, 1 OA book chapter, 9 OA conference proceedings, 10 additional publications in non-OA venues (due to publisher restrictions). Furthermore, 6 OA journal articles are under revision, and 5 more are in preparation.
The consortium actively engaged with the scientific community, participating in 43 conferences and 14 workshops, and organizing 8 additional workshops to share findings. Outreach efforts extended to the general public through exhibitions, radio interviews, festivals, media coverage, and newspaper articles.
In terms of intellectual property and commercialization, 1 patent has been granted and 3 more filed and 2 new 3D printing materials were developed and are now commercialized by the SME partner in the consortium.
Bioinspired control strategies were also explored, drawing on stereotypical elephant movements. These strategies leverage environmental interaction and tactile feedback to provide the robot with easily learnable models, bodily awareness, and internal representations of its surroundings. This work supports the development of intuitive, adaptive robotic behavior.
The project contributed to additive manufacturing through the creation of new functional material formulations and advanced production techniques, resulting in commercialized products.
The artificial trunk’s versatile grasping capabilities pave the way for novel applications in both academic and industrial research focused on dexterous robotic manipulation. In the long term, PROBOSCIS has the potential to evolve into a ‘living’ assistant, easing human-robot interaction in challenging contexts. It could assist with tasks such as aiding mobility, lifting loads, or interacting gently with patients—especially valuable in healthcare and assistive environments. In search and rescue operations (e.g. after earthquakes or hurricanes), the robot could carefully remove debris and identify victims even in low-visibility conditions like darkness or smoke, with sufficient strength to safely move people to secure locations.
Industrial applications are equally promising. In the food industry, PROBOSCIS could handle diverse items and assess their condition without relying on vision, particularly when objects are occluded during manipulation. It also offers solutions for complex tasks like picking and sorting objects of various shapes and materials—highly demanded in e-commerce, warehousing, and logistics—or manipulating deformable materials such as paper or fabric.
The project resulted in the publication of: 27 papers in open-access (OA) journals, 1 OA book chapter, 9 OA conference proceedings, 10 additional publications in non-OA venues (due to publisher restrictions). Furthermore, 6 OA journal articles are under revision, and 5 more are in preparation.
The consortium actively engaged with the scientific community, participating in 43 conferences and 14 workshops, and organizing 8 additional workshops to share findings. Outreach efforts extended to the general public through exhibitions, radio interviews, festivals, media coverage, and newspaper articles.
In terms of intellectual property and commercialization, 1 patent has been granted and 3 more filed and 2 new 3D printing materials were developed and are now commercialized by the SME partner in the consortium.
PROBOSCIS is projected to play a pivotal role in advancing science and technology, as researchers have uncovered fundamental principles of elephant trunk behavior that can inspire novel grasping and manipulation technologies. For the first time, biological and behavioral studies of the elephant trunk revealed entirely new scientific findings, including previously undocumented anatomical structures. This knowledge enabled the creation of a 3D virtual reality model of the trunk for educational and outreach purposes.
Beyond its technological relevance, the research on trunk behavior is expected to support conservation efforts by raising awareness about the species and informing strategies for safe human-animal interactions and conflict avoidance.
The project’s main breakthrough lies in the bioinspired design of a continuum robotic manipulator that eliminates the traditional distinction between arm and gripper, mimicking the elephant trunk’s full-arm versatility in grasping. This innovation led to the development of new biomimetic artificial muscles, robust and distributed soft tactile sensing systems, and advanced strategies for component integration.
Equally critical are the bioinspired control strategies being developed—such as motion patterns modeled on stereotypical elephant movements and environmental interaction—enabling robots to acquire easily learnable models, bodily awareness, and internal representations of their surroundings.
Advances in materials science are central to the project. The development of tough yet stretchable functional materials for sensing and actuation is ongoing, along with fabrication processes suitable for producing large-scale flexible components. These innovations will benefit additive manufacturing industries, extending the applicability of new material formulations across various sectors.
The versatile grasping capabilities of PROBOSCIS are expected to open new frontiers in dexterous robotic manipulation for both academic research and industrial applications. In the long term, PROBOSCIS is envisioned as a ‘living’ assistant capable of enhancing human-robot interaction, especially in challenging environments. For instance, it could aid in search and rescue operations by clearing debris to detect victims—functioning even in darkness or smoke without visual input—and possess sufficient strength to safely lift and move people to safety.
In industrial settings, PROBOSCIS could be deployed in the food sector to handle a wide range of items and assess their state without relying on vision. It also holds promise for e-commerce and warehouse automation, enabling the picking and sorting of multiple objects of varying shapes and materials. Additionally, it offers a solution for manipulating deformable items such as paper or fabrics, tasks that remain particularly challenging for conventional robotic systems.
Beyond its technological relevance, the research on trunk behavior is expected to support conservation efforts by raising awareness about the species and informing strategies for safe human-animal interactions and conflict avoidance.
The project’s main breakthrough lies in the bioinspired design of a continuum robotic manipulator that eliminates the traditional distinction between arm and gripper, mimicking the elephant trunk’s full-arm versatility in grasping. This innovation led to the development of new biomimetic artificial muscles, robust and distributed soft tactile sensing systems, and advanced strategies for component integration.
Equally critical are the bioinspired control strategies being developed—such as motion patterns modeled on stereotypical elephant movements and environmental interaction—enabling robots to acquire easily learnable models, bodily awareness, and internal representations of their surroundings.
Advances in materials science are central to the project. The development of tough yet stretchable functional materials for sensing and actuation is ongoing, along with fabrication processes suitable for producing large-scale flexible components. These innovations will benefit additive manufacturing industries, extending the applicability of new material formulations across various sectors.
The versatile grasping capabilities of PROBOSCIS are expected to open new frontiers in dexterous robotic manipulation for both academic research and industrial applications. In the long term, PROBOSCIS is envisioned as a ‘living’ assistant capable of enhancing human-robot interaction, especially in challenging environments. For instance, it could aid in search and rescue operations by clearing debris to detect victims—functioning even in darkness or smoke without visual input—and possess sufficient strength to safely lift and move people to safety.
In industrial settings, PROBOSCIS could be deployed in the food sector to handle a wide range of items and assess their state without relying on vision. It also holds promise for e-commerce and warehouse automation, enabling the picking and sorting of multiple objects of varying shapes and materials. Additionally, it offers a solution for manipulating deformable items such as paper or fabrics, tasks that remain particularly challenging for conventional robotic systems.