Final Report Summary - MOONWALK (Technologies and Human-Robot Collaboration for Surface EVA Exploration Activities and Training in European Analogue Environments)
Executive Summary:
MOONWALK (http://www.projectmoonwalk.net) developed innovative technologies for human-robot interaction and co-operation in Extra Vehicluar Activities (EVAs) in the context of the exploration of Moon and Mars (robot-assisted EVA). The MOONWALK scenario foresees one or two astronauts equipped with EVA suits (e.g. space suits) and special tools (e.g. for exobiology sampling) that are supported by a small, all-terrain assistant-robot (rover).
In addition to the technology development, MOONWALK built capacities and know-how related to European infrastructures for Earth analogue simulations. Such infrastructures are needed to validate European Space technologies under realistic environmental conditions while still being affordable.
The main technical achievements of MOONWALK were the development of core components needed for a robot-assisted EVA mission, including a simulation EVA space suit, a small helper robot, a user-friendly man-machine-interface (MMI), EVA information- and communication system system, a number of manual and robot-mounted tools to collect scientific samples, and a device to improve the online monitoring of the astronaut's vital signs.
The MMI was of particular interest, as it is needed to enable an astronaut in a heavy space suit to efficiently interact with the robot. The space suit severely restricts the mobility, the dextrosity, and the visual prerception capabilities of the astronaut. To overcome these restrictions, the MMI uses manual gestures for robot control and a data interface integrated in the EVA suit.
To demonstrate and validate the project results, two earth analogue simulations were conducted in the last year of the project. The first one was in Rio Tinto, Spain. In this campaign, a robot-assisted EVA on Mars was simulated. Rio Tinto has a morphology and a soil-chemistry that can be compared to Mars. Consequently, the focus of the mission was on exploration of the terrain and gathering of soil samples. Through a co-operation with the EU-funded SHEE project and the integration of the SHEE habitat in the simulation, a realistic EVA mission, including egress and ingress of the astronaut to a planetary habitat, could be simulated. The second analogue simulation was off the coast of Marseille, France. Here the mission scenario included the exploration of the lunar surface under simulated low-gravity conditions.
From the start, MOONWALK put a strong focus on public outreach and communication with European and international stakeholders, including ESA and NASA. Selected by a public Announcement of Opportunity, several experiments contributed by external stakeholders were integrated in the Rio Tinto and Marseille missions. Observers from ESA and NASA, including a real ESA astronaut, attended the simulations and provided valuable feedback. Through a children's competition for the best quote from the first astronaut on Mars and the best design for a Moon mission flag, the interest of the young generation in Space exploration and human spaceflight was ignited. Overall, the massive public and media interest in the MOONWALK simulations is proof that Space exploration is still a fascinating topic and that earth analogue simulations in Europe are not only valuable to validate key technologies, but also to efficiently promote human space flight .
Project Context and Objectives:
Today, the exploration of Moon, Mars and other objects in Space, such as asteroids, is done exclusively with the help of robots and automated probes. NASA’s Curiosity mission is a recent example of a successful robotic exploration mission to Mars. With ExoMars, ESA has plans for a European robotic mission to Mars in 2020, and the Chinese sent a robot to the Moon in 2013.
However, the full potential of Space exploration can only be reached with human astronauts on-site, and thus both NASA and ESA are working on plans for human missions to Mars, Moon and asteroids in the years after 2020. In these missions, humans will be accompanied and supported by robots, and a close co-operation of humans and robots will be required.
The objective of the MOONWALK project was the development and testing of technologies for human-robot interaction during Extra Vehicular Activities (EVA) on planetary surfaces. The control of robots on a planetary surface is non-trivial because astronauts are limited in their movements by a bulky spacesuit and the effects of reduced gravity. MOONWALK developed new, practical methods for the control of robots by astronauts. These methods were tested in two earth-analogue simulations in a desert near Rio Tinto (Spain) and underwater off the coast of Marseille (France). In both cases, a planetary EVA was simulated under environmental conditions similar to those found on Mars (e.g. acidic soil chemistry, rugged terrain, dust contamination, communication delay) and the Moon (e.g. reduced gravity).
In both trials, an astronaut was made to wear a simulation spacesuit that imitated the properties of a real pressurized spacesuit, like weight, movement constraints, and visual constraints. The astronaut was provided with a small helper robot and newly developed interfaces and tools to enable robot control, communication with the local mission control (e.g. in the lander) and the mission control center on Earth, and on-site soil sampling and science.
Seven beneficiaries from seven EU member states participated in MOONWALK: The DFKI Robotics Innovation Center in Bremen (Project Coordinator), COMEX in France (technical coordination), EADS Airbus Defense Systems in the UK, LIQUIFER Systems Group in Austria, Space Application Services in Belgium, NTNU Centre for Interdisciplinary Research in Space in Norway, and the Instituto Nacional de Técnica Aeroespacial (INTA) in Spain. Intensive contact to external stakeholders, such as ESA and NASA, were established during the project.
The overall objectives of the MOONWALK project can be summarized as follows:
- Develop innovative technologies to support co-operative human-robot exploration of planetary surfaces (robot-assisted EVA);
- Use and further develop infrastructures, technical systems, and know-how needed for analogue simulations in Europe;
- Test and demonstrate the technologies developed in MOONWALK in two analogue simulations (in Rio Tinto and Sub-Sea Marseille) under simulated Mars and lunar conditions;
- Reach out to the (European and international) Space community and to the public to boost the interest in human space flight and in European analogue simulations.
The following is a list of the main MOONWALK objectives and their requirements, and how these objectives were addressed and met in the MOONWALK project:
O1 requirements: Human-robot and human-human cooperation in extreme environments with shared robot control between the Control Centre and the astronaut on-site. This includes wearable Human-Machine Interfaces (HMI) that work in extreme environments and under difficult operation conditions.
O1 achievement: Various human-machine-interfaces (wrist display, tablet, gesture control) and a control center were successfully implemented and tested in relevant environments (EVA simulations). Proof-of-concept that the interfaces work under difficult conditions was provided, albeit at a low TRL and without full integration of direct robot control through.
O2 requirements: Adaptation of an (existing) autonomous operating rover-type robot platform for human-controlled interaction inclusive required instrumentation (cameras, sensors, communication and navigation package).
O2 achievement: The YEMO helper robot is a new construction based on the DFKI ASGUARD system, fulfilling the requirements of MOONWALK (amphibious operation in difficult terrain). Sensors integrated in YEMO included cameras and a communication package. Because the focus of MOONWALK was on astronaut-robot interaction and control and the budget was limited, the final decisional autonomy of the robot was low and limited to a follow-me functionality.
O3 requirements: Design and setup of communication, mission planning and operations infrastructure, as well as tools (control centre) which can be adapted to various mission scenarios (such as Moon or Mars with variable communication delays).
O3 achievement: A mission control infrastructure was set up with an on-site control center simulating CapCom, or the local control center on the planetary surface, and a remote control center in Brussels, simulating mission control (MCC) on earth. The set-up was successfully tested in both analogue simulations. Variable communication delays between MCC and CapCom were implemented.
O4 requirements: Evaluate human performance in extreme environments (in function of gravity level variations, pressure and temperature) and the estimate the effect of those environments on robot-human cooperation. The goal was to establish lessons learned for the design of future exploration missions and associated technology, and to identify the factors and stressors which affect human performance and to develop mitigation strategies.
O4 achievement: Two earth analogue simulations were performed in a desert-like environment (Rio Tinto, Spain) and underwater off the coast of Marseille to simulate extreme environments as much as possible within the framework of the MOONWALK project. Human performance (of mock-up astronauts) was measured during the analogue simulations, both with technical means (bio-monitoring) and with empirical means (questionnaires). Lessons learned were extracted and formulated in deliverables D8.1 and D8.2. Although these results prove that robots can be valuable helpers for humans even in extreme environments, the overall scope of the analogue simulations possible in MOONWALK was too limited to support a scientifically sound empirical assessment to identify stressors and develop mitigation strategies.
O5 requirements: Improve the sustainability of life in extreme environments through protection garment and portable life support systems for astronauts. The devices tested in MOONWALK should include a bio-monitoring system integrated in the EVA space suit to monitor basic life signs of the astronaut during the EVA. Physiological correlations between crew activity, life support suit performance, crew health and subjective well-being of the mock-up astronauts should be established.
O5 achievement: A new EVA simulation space suit that can be used for both terrestrial and underwater simulations was designed and build in MOONWALK. For the simulations, a bio-monitoring system was integrated in the EVA suit. The collected data was used to monitor the health performance of the astronauts during the analogue simulations (see D8.1 and D8.2).
O6 requirements: Define new robot-supported search methodologies and strategies to detect extremophile life forms and bio-signatures in terrestrial analogues by integrating existing hardware in the mission scenarios.
O6 achievement: Devices for in-situ analysis of soil probes, a RAMAN spectroscope and the SOLID in-situ lab tool were customized to be carried by the YEMO helper robot. The tools and handling by the helper robot could prove the general functionality and usability of the concept. However, due to technical problems during the simulations, the full potential of the robot-mounted tools could not be shown.
O7 requirements: Develop manual and robotic sampling tools and field exploration procedures and test them in extreme environments (low temperature, microgravity) and for different application fields (geology and exobiology). Assess the effectiveness of the tools when used by astronauts in extreme environments.
O7 achievement: Several manual sampling tools and robot-mounted support tools were developed and implemented. A set of field exploration procedures was defined and used to test the tools in the analogue simulations. The usability of the tools by the astronauts was evaluated empirically. Unfortunately an active robotic sampling tool (e.g. manipulator arm) could not be implemented due to budget constraints, which limited the usability of the robot for sample collection. Also, a scientifically sound quantitative evaluation of the efficiency gains created by the manual tools could not be provided due to the limited scope of the analogue simulations.
Project Results:
The main results of MOONWALK were A.) technologies and tools, B.) new capacities related to European infrastructures for earth-analogue simulations, and C.) data on the feasibility of human-robot interaction on extraterrestrial planetary surfaces.
A.) With respect to technologies and tools, the main results were:
EVA simulation space suit: COMEX developed a concept and a design for a new EVA simulation and training suit. The design was influenced by previous work and experience of COMEX with the GANDOLFI EVA training suit. However, efforts were made to take also new designs of NASA suits into account: The new training suit was based on the same architecture and anthropometric capabilities of the NASA Z-series spacesuits prototypes and includes a rear entry system as the Russian ORLAN suit. In addition, the concept took into account the requirements for robot-assisted EVA identified in MOONWALK, the results of human factor analysis, and the operational requirements of the planned Earth Analogue Simulations.
The EVA training suit incorporates an exoskeleton made of hard and soft parts, and a diving gear for the astronaut diver for the underwater configuration. The exoskeleton design had three main drivers, to take into account ergonomics aspects of the Z-1 suit (NASA input to MOONWALK), to simulate constraints of movements as in a real pressurized space suit, and to put the astronaut/diver into neutral buoyancy (or reduced gravity). This function is only available in underwater configuration.
The exoskeleton is composed of hard shells made of composite materials that enable a significant weight gain compared to standard metal materials with a high mechanical resistance. Those parts are linked with neoprene parts.
The second major hard elements are the bearings allowing rotations of the different limbs, whom rings are made of POM and ball made of 316L steel (AISI standard).The bearings location is directly inspired by the Z1 series American spacesuit to allow rotations of shoulders, arms, thighs and pelvis.
These hard parts are linked through soft neoprene parts, tailored-made manufactured by Beuchat company, fixed on bearings.
To simulate the movement-constraints caused by a pressurized spacesuit the movement amplitudes of the training suit were limited by the design itself to comply with the Z1 spacesuit movement constraint specifications. In addition, a mechanical resistance was induced on the joint locations. The joint elements on the leg and arm assemblies (knee and elbow) restrain movements by creating a resistant torque. This makes the flexion of the members harder. The joint elements are torsion springs, which are sized to be the closest of the Z1 spacesuit movement resistances.
Astronaut subjects using the EVA suit for MOONWALK are limited to the following heights: minimal 160cm to maximal 195cm. To adjust arm lengths and leg lengths, adjustment tabs are located on both elbow and knee joints as shown below. To adjust the torso-pelvis distance, the length strap located on the torso can be changed according to the subject’s height.
The retractable visor is made of PMMA (ALTUGLAS®) and fixed on the helmet authorizing a rotation for allowing it to be used in down or up position. The field of vision has been optimized through the helmet design and the full face mask performance and can theoretically reach 60°.
The design of the gloves for the space suit was based on current NASA EVA gloves. The gloves were manufactured from a neoprene classic diver gloves recovered by the outer fabric of standard ski gloves. Home-made silicon parts were manufactured and stuck on to the gloves outer skin in order to make them look like NASA gloves.
Human-Machine Interface (HMI) and EVA Information System: The main features of the HMI were a chest display and a wrist display, both developed by Space Application Services. The chest display consisted of a tablet computer which is located on the torso of the space suit. The touch-screen of this device was modified to work underwater. The interface displayed on the chest display was set up to enable the use with heavy space gloves. The chest display was mainly used to provide information related to the mission procedures (recorded in standard NASA format) to the astronaut.
The wrist display was intended a.) to replace standard US EVA cuff checklist on the spacesuit and b.) to provide an alternative to the gesture control (see below) of the robot. This display used a small a screen to display an interface consisting of an array of simplified push buttons.
A Suit Computer Assembly (SCA) was integrated in the space suit to enable communication between the astronaut and the local mission control center. The SCA was a computer unit in a waterproof housing that does the necessary processing for the Human Machine Interfaces, but also for other potentially required sensors, including the Rover Gesture Control. In the underwater simulations, the SCA connected to a communication buoy via an umbilical. In the Rio Tinto simulations, an wifi connection was used to connect to CapCom.
A Mission Control Centre was set up at Space Applications in Brussels, Belgium. During the simulations, this control center simulated the MCC on earth. Consequently, a time-lag of up to serveral minutes between MCC and CapCom was implemented to test the effects of delayed response and local vs. remote decision making..
Robot control-by-gesture: In addition to “traditional” HMI solutions, MOONWALK investigated the feasibility of an innovative way to interact with robots. The approach developed by DFKI uses simple gestures made by the astronaut with his/her upper limbs. The gestures are recorded with IMUs (Inertial Measurement Units) embedded in the space suit, interpreted with a special software, and then translated to control commands for the robot. Using IMUs instead of optical sensor (cameras, 3-D cameras), which is a more common approach, has several advantages:
- The gesture recording is independent of variable visibility conditions. This is important for using the system in earth analogue simulations underwater, but will also play a significant role in a potential application in future Mars missions;
- The IMU’s are water-proof and can thus be used in underwater analogue simulations;
- The astronaut does not have to be in the line-of-sight of the robot, i.e. the astronaut can control the robot independent of his/her relative position to the robot and also record the pose of the astronaut in emergency situations (i.e. astronaut lying on the ground).
In MOONWALK, a simple set of gestures was implemented and tested, both in the lab and in both earth analogue simulations. The method proved to generally work very well and, after some initial training, was described by the astronauts to be easy to use and practical. However, improvements in the reliability and performance of the underlying software are still necessary. For the application of the gesture control in terrestrial spin-out applications (e.g. for the control of industrial robots), the enlargement of the gesture repertoire is planned.
Amphibious helper robot: DFKI used the general design of their well-tested and proven ASGUARD robot family as the basis for the MOONWALK robotic helper. This concept was modified to accommodate the requirement of a system that can operate both on land and underwater. The result, the amphibious YEMO rover, featured a water- and pressure-proof (up to a water depth of 10 m) body with four individually actuated star-wheels or regular wheels. The weight of the robot under terrestrial conditions was around 15 kg, but could be adjusted to simulate lunar conditions underwater. The size of the robot is approx. 1,5 m x 0,4 m. With the four star wheels, this small rover can climb steep slopes and handle difficult terrain.
The main sensor integrated in the rover was a watertight 360 degree camera. With this camera, a constant stream of 360 degree images can be created and analyzed. The set-up enables several analysts in the mission control center to analyze different parts (e.g. view angles) of the data in parallel.
A payload box developed by Liquifer Systems Group was mounted on the rover. It contained the manual tools and was designed to hold soil and rock samples. It also contained the RAMAN spectrometer, which was modified to fit on the robot and in the payload box.
Biomonitoring system: The underlying principle of the Biomonitoring system developed by EADS / Airbus was that the astronaut’s physiological signs can be monitored to improve the overall picture of their welfare. This is achieved through several systems: the monitoring of the user’s heart rate (through traditional beats per minute (BPM) recording) and the assessment of the physiological and mental stresses being experienced by subjects.
The system uses a machine learning approach with classifiers to determine, from real life data, the physical state of the astronaut. As this is a statistical approach, there is not a requirement to understand the underlying physiological nature of the readings you are gathering, but instead to recognize patterns and behaviors in the underlying data stream and how they relate to previously acquired data.
In order to build this system, appropriate statistical features had to be identified, which was a key focus for the research in MOONWALK. Prior research on the subject, and the experience of the research team, has focused on Heart Rate Variability (HRV) as the primary source for these features. HRV is the change in time between each heartbeat, and is affected by physiological and mental stress.
Many of the candidate features are focused on the frequency domain, understanding the various components that make up the overall ‘signal’ of the Heart Rate Variability. These features can be fused together, such as ratios between high frequency and low frequency components.
After identification of appropriate features, the classification framework was ‘trained’ with test data acquired from various subjects in different conditions of physiological and mental stress, alongside a ‘ground truth’ to represent an indication of the stress levels experienced.
The current prototype system classifies the user’s stress into three broad categories: Red, Amber, Green. This ‘traffic light’ system acts as a general indicator, ‘Green’ being a relaxed, non-stressed state, ‘Amber’ indicating medium physical and mental stresses, and ‘Red’ meaning a high level of physical and cognitive loading.
Tools for manual, robot-assisted sampling: The robot operational scenarios developed in MOONWALK comprise a range of astronaut-robot interactions of various degrees of complexity. Liquifer Systems Group developed a number of manually operated tools to support the astronaut. In particular, these were an Astronaut Rescue Tool (ART), an Astronaut Tether Control (ATC), a Pantograph Sampling Tool (PST), a Foldable pick-up Claw (FPC), and a Manual Tool Rack (MTR). All the manual tools aim at safe operation through a single astronaut in cooperation with a robotic rover.
The Astronaut Rescue Tool (ART) for astronaut assistance should he/she fall over is a manual tool which is positioned in a separate compartment of the Payload Box (PB) installed on the robot. The collapsible tool is single-handed accessible to the fallen astronaut and can be deployed single-handed by flicking the arm. It is developed from an off-the-shelf walking stick with adapted handle and anti-slip elements along the stick. The Astronaut Tether Control (ATC) is the manual control for the rover scouting a steep slope. ATC is developed from an off-the-shelf retractable dog leash with adapted control buttons and modified handle to fit the astronaut glove. The end of the tether is modified for easy attachment to the rear of the rover.
The Pantograph Sampling Tool (PST) is a retractable sampling tool which can be used single handed to collect samples and feed them into sampling bags. The design is being developed from a lazy tong concept which allows elongation and contraction of the tool arm. It is equipped with a scoop gripper. The Foldable pick-up Claw (FPC) for sample collection is based on an off-the-shelf foldable system. Both collapsible tools can be transported in the payload box of the robotic rover or can be attached to the astronaut suit when feasible. The Manual Tool Rack (MTR) replaces the manual tool cart which was foreseen for (sampling) tool transport in earlier scenario versions. The rack simulates a rack for bigger existing tools (supplied by COMEX), sampling bags and construction material similar to the Lunar Roving Vehicle (LRV) rack. It can also serve as support to secure the Astronaut-rover tether operations with a winch system.
Using rapid prototyping and 3-D printing, workable prototypes of the manual tools were developed and tested in the earth analogue simulations in Rio Tinto and Marseille.
B.) With respect to the earth analogue simulations, the main results were:
Rio-Tinto Simulation: The technical equipment, methods and mission procedures developed in MOONWALK were verified in a Mars analogue simulation in Rio Tinto (Spain) between April 16th an 30th 2016. A planetary exploration mission was simulated in which a team of an astronaut and a robot executed different tasks and scenarios required to achieve the mission objectives: i) Find a safe and secure site for human settlement; ii) Finding resources: minerals, material for contractions (e.g. gravel, rocks); iii) Astrobiological research by searching for signs of live. The performance of the astronaut-robot team was compared qualitatively to the performance of an astronaut-astronaut team doing the same EVA tasks.
The whole campaign lasted two weeks. The first week was devoted to setting up, testing and training all the equipment and procedures. The planetary exploration mission simulation was performed during the second week. The main elements for the simulations were: a prototype of a Martian habitat (SHEE habitat), the Gandolfi II astronaut training suit, a small helper robot (YEMO) gesture-controlled by the astronaut and equipped with different tools developed during MOONWALK, and scientific instrumentation for astrobiological investigations.
An average of 25 persons from the different partners of the MOONWALK consortium participated directly in the different aspects of the campaign. Among them, a medical team for medical check of the simulation astronauts, as well as the participants in the external experiments.
The main achievements of the campaign were: i) Multiple EVA simulations involving only astronauts or hybrid astronaut-robot teams were carried out to test and execute the tools, procedures and scenarios developed during the MOONWALK project, including gesture-controlled astronaut-robot scouting, sampling, or exploring inaccessible sites by any of them separately; ii) A realistic Mars exploration mission simulation was performed with the main objectives achieved: the landing site was explored and mapped, different resources were identified, and microbial markers (evidences of life) were detected in the collected samples; iii) extensive outreach and dissemination of the campaign both in the field and through press, TV, radio, both national and international (See D9.6).
Marseille simulation: The simulation campaign in Marseille has been set up accordingly to one of the MOONWALK project’s task definitions: “The objective is to perform a simulated Moon mission at an analogue site offshore Marseilles. The objective of these tests will be to validate the human-robot interaction architecture and to simulate various activities that future astronauts might perform on the Moon (cave exploration by robot, crater descent, installation of instruments and mock-up habitation structures. “
Nine potential lunar analogue sites were identified in the area of Marseilles, in the Calanques National Park. Situated at a reasonable water depth, they were selected mainly for their geomorphological similarity with some interesting spots on the lunar surface. Among those sites Port de Pomègues was selected for the MOONWALK lunar surface EVA simulations. The reasons for that choice were:
1. The site hosts several geological features of interest, namely a crater-like structure (which in reality is a sinkhole) and steep edges on the borders of the site.
2. Its exposition to wind is low (which is quite important for the Marseilles area where strong winds can quickly block such tests for several days) and the operations vessel MINIBEX can navigate in the zone.
3. A preliminary survey on the strength of communication signals showed that this site is best covered amongst the nine sites identified.
The underwater analogue simulations in Marseille were conducted between May 30th 2016 and June 10th 2016. Because of the complexity of the simulation architecture and of diversity of equipment involved and managed by the different partners, preliminary EVA pool tests were necessary to train the divers who were selected to be astronauts and experience for subsea operations (particularly for security procedures rehearsal). Thus the week from the 30th of May to the 3rd of June was dedicated to pool trials in the COMEX’ facilities. The second week from the 6th of June to the 10th of June was devoted to the subsea analogue simulation to Moon with the purpose of testing the Astronaut-Robot and Astronaut-Astronaut procedures.
Human factor evaluation: In MOONWALK, analogue simulations we evaluated the human performance and the effect of robot-human cooperation in very different environments and in interaction with different simulation components, i.e. astronaut suit, astronaut tools, human machine interfaces (HMI). The simulations generated several types of data; video and audio recordings, GPS tracking, biomonitoring, written logs and questionnaires and interviews. However, due to the considerable amount of data, technical challenges and time limitations the human factor evaluation are limited to data and information from written logs (CAPCOM), questionnaires and interviews.
The findings of the trials indicate that the different sites lead to differences in how the astronauts evaluate the system components and their experience of the simulations. However, there is little indication that the interaction with the rover/robot resulted in different responses from the simulation astronauts compared to interaction with another astronaut, i.e. the robot was seen by the astronauts as a valuable and useful helper.
Generally, simulations in analogue sites have shown to represent great tools to learn about the complex socio-technical system that a human exploration mission will represent. However, when comparing terrestrial (Rio Tinto) versus underwater (Marseille) simulations, the main conclusions are that
- terrestrial simulations are a very useful mean to work on the technical aspects of a complex simulation architecture. It allows engineers easily to stand next to the simulation-astronauts, try out tools, operations and procedures and change those based on the feedback from the field. Test can be quickly performed and adapted. Terrestrial simulations are therefore very useful to improve technologies and do engineering on EVA equipment.
- subsea simulations are less useful for such engineering considerations (or “debugging”). Once all the materials is on the ship or underwater all has to work. Delays and small errors lead to large delays in the operations since various security issues have to be respected. However, from feedback from the subjects that served in both simulations a clear feedback was that the subsea simulation is much more realistic in terms of technical difficulty and mental stress. Subsea simulations are therefore very useful for training purpose of astronauts.
Conclusions:
- All the MOONWALK elements were fully deployed and have been fully operating: the robot YEMO, the training EVA suit Gandolfi II, the tools for collecting and sampling, the biomonitoring system, and the communication system.
- This first attempt of a European subsea analogue simulation with a robot went beyond the proof of concept, and had shown the relevance of team/robot cooperation through the realistic procedures scenarios. Invited visitors from ESA confirmed this interest in such simulations in Europe in the future.
- The outreach was extremely successful with a world-wide coverage of still photography and video material.
Potential Impact:
Impact: MOONWALK developed and evaluated new technologies for human-robot interaction, built know-how and equipment in support of European capabilities for earth analogue simulations, and raised the awareness for human spaceflight and the benefits and opportunities of earth analogue simulations with European and international stakeholders as well as with the general public.
On a political level, MOONWALK helped to create an understanding for the necessity to build and use European infrastructures for earth analogue simulations to test space technologies and prepare for future missions with a justifiable effort. The project also helped to raise the public awareness of space exploration and to re-ignite the fascination in human space flight, in particular in the young generation.
Overall, MOONWALK was a showcase project which not only inspired people to make the next step into human robotic space exploration but actually already connected to one of the possible end costumers, ESA, for the inclusion of MOONWALK products into their studies and programmes.
Exploitation: On a technical level, the project developed and tested a number of technologies and equipment that will be available for future earth analogue simulations and be scientifically and commercially exploited by the project partners for space-related and terrestrial applications.
The general MOONWALK exploitation foresees that all partners are allowed and encouraged to exploit, both commercially and non-commercially, any technical results, know-how, or facilities they developed during the MOONWALK project as they wish. Legally, any exploitation activity has to conform with the rules for the distribution of intellectual property rights (IPR) as defined in the Consortium Agreement (CA). These relevant paragraphs of the CA are listed in Appendix A.
Overall, the CA adopts a policy that gives IPR on knowledge and/or solutions (technologies) to those partners that developed the knowledge / solutions within the project. Where several partners worked on the same exploitable results together during the project, specific conditions for sharing these results (licenses) have to be negotiated between the partners.
In terms of exploitation of project results by individual consortium members, there are generally two different types of organizations that use different general exploitation strategies:
• University and Research Partners have a focus on the non-commercial exploitation of project results. The primary exploitation channels for these partners are to
o use project results in on-going national and European R&D projects;
o leverage project results, contacts and expertise gained in the project to apply for new R&D projects;
o to include project results in the graduate (and undergraduate) teaching;
o to use scientific project results for on-going or new PhD projects;
o to market high-TRL results through spin-off companies (if feasible as joint-ventures with other project partners).
• Industrial Partners are mainly interested in the commercial exploitation of project results. Their main channels for exploitation are related to a
o further evaluation and progressive prototyping of the promising project results;
o integration of specific technologies developed in the project in existing products;
o the “spinning off” of new products.
Almost all of the technical outcomes described in the previous section can be applied in terrestrial applications. An example is the control-by-gesture, a method that is already exploited by DFKI in support of applications related to Digital Industry (i.e. robot-control in an industrial environment). Another example is the bio-monitoring developed by EADS / Airbus which is used for operators in extreme environments.
Other outcomes will be exploited in future space-related projects and applications. An example is the simulation space-suit Gandolfi II which will be offered by COMEX to future earth analogue simulations as part of R&D projects or campaigns preparing for future space missions.
Dissemination: The outreach and dissemination goals of MOONWALK were to establish a high visibility of the project not only within the scientific research community, but also on a much broader scale that includes the general public, the (space) industry and the stakeholders (including European and international space agencies).
The idea was to educate and enlighten the target audience about the possibilities of working with machinery in extreme environments, highlight applications of space technologies on earth and attract future research, academic and industrial partners for further development (including commercialization) of the concepts and prototypes created through MOONWALK.
As outlined in Final Dissemination Report (D9.6) the consortium partners under the lead of Liquifer Systems Group have used a variety of means and tools to achieve these goals. This included a well-designed and maintained web site, four public newsletters, flyers and fact-sheets, and an extensive project video summarizing the approach and results of the analogue simulations.
The project was presented with scientific conference papers at various international conferences, including IAC 2014 and 2015, IAASS 2014, ASTech 2014, Astrobiology Science Conference 2015, AIAA Space 2015, ESREL 2015, International Symposium on the Moon 2020-2030. In addition, papers in Astrobiological and Space-related journals were submitted. There was significant media coverage of the project and the analogue simulations, with more than 50 press articles and TV reports (see D9.6 Final Dissemination Report for details).
In addition, the project managed to create significant synergies with other EU-FP 7 projects. An example was the SHEE (Self-deployable Habitat for Extreme Environments) project (http://www.shee.eu/main). The planetary habitat test-bed for terrestrial analogue simulations was used at the MOONWALK simulations in Rio Tinto for an astrobiology laboratory and the Local Mission Control. Specifically for the MOONWALK project a suitport was built to the SHEE habitat to simulate ingress and egress procedures for the first time in Europe.
Dissemination of MOONWALK towards the European and international space agencies was high on the agenda of the MOONWALK team. MOONWALK connected and cooperated with NASA for the exchange of ideas and has enabled a feedback loop that has proved useful to both. There has also been a strong tie to ESA with regard to Moonwalk components, for example the suit, the tools, the communication architecture amongst other parts. There are two study projects with ESA, one was called “LUNA Analogues for Preparing Robotic and Human Exploration on the Moon – Needs and Concepts” [RD12], the other one is called MOONDIVE (ongoing). Both are dedicated to artificial analogues in preparation of future lunar exploration as a described by the ESA DG Jan Wörner in his Moon Village concept. In both concepts MOONWALK hardware is integrated and become key components of the ESA concepts. MOONDIVE is a study about scenarios for the Neutral Buoyancy Facility at ESA-EAC with the Gandolfi 2 suit to train astronauts for future lunar exploration. LUNA is about a whole artificial analogue to be established at ESA-EAC to support astronauts in training for future lunar or Martian surface missions. Furthermore, MOONWALK was introduced at the ESA Isolation Steering Committee meetings as a facility ready to be used for future ESA isolation studies.
MOONWALK supported a number of strategic activities, including the EC Analogue Workshop held in Brussels on October 12th 2016.
List of Websites:
MOONWALK Website: http://www.projectmoonwalk.net/
Coordinator contact: Dr. Thomas Vögele, DFKI Robotics Innovation Center, Robert-Hooke Str. 1, 28359 Bremen (Germany), thomas.voegele@dfki.de
MOONWALK (http://www.projectmoonwalk.net) developed innovative technologies for human-robot interaction and co-operation in Extra Vehicluar Activities (EVAs) in the context of the exploration of Moon and Mars (robot-assisted EVA). The MOONWALK scenario foresees one or two astronauts equipped with EVA suits (e.g. space suits) and special tools (e.g. for exobiology sampling) that are supported by a small, all-terrain assistant-robot (rover).
In addition to the technology development, MOONWALK built capacities and know-how related to European infrastructures for Earth analogue simulations. Such infrastructures are needed to validate European Space technologies under realistic environmental conditions while still being affordable.
The main technical achievements of MOONWALK were the development of core components needed for a robot-assisted EVA mission, including a simulation EVA space suit, a small helper robot, a user-friendly man-machine-interface (MMI), EVA information- and communication system system, a number of manual and robot-mounted tools to collect scientific samples, and a device to improve the online monitoring of the astronaut's vital signs.
The MMI was of particular interest, as it is needed to enable an astronaut in a heavy space suit to efficiently interact with the robot. The space suit severely restricts the mobility, the dextrosity, and the visual prerception capabilities of the astronaut. To overcome these restrictions, the MMI uses manual gestures for robot control and a data interface integrated in the EVA suit.
To demonstrate and validate the project results, two earth analogue simulations were conducted in the last year of the project. The first one was in Rio Tinto, Spain. In this campaign, a robot-assisted EVA on Mars was simulated. Rio Tinto has a morphology and a soil-chemistry that can be compared to Mars. Consequently, the focus of the mission was on exploration of the terrain and gathering of soil samples. Through a co-operation with the EU-funded SHEE project and the integration of the SHEE habitat in the simulation, a realistic EVA mission, including egress and ingress of the astronaut to a planetary habitat, could be simulated. The second analogue simulation was off the coast of Marseille, France. Here the mission scenario included the exploration of the lunar surface under simulated low-gravity conditions.
From the start, MOONWALK put a strong focus on public outreach and communication with European and international stakeholders, including ESA and NASA. Selected by a public Announcement of Opportunity, several experiments contributed by external stakeholders were integrated in the Rio Tinto and Marseille missions. Observers from ESA and NASA, including a real ESA astronaut, attended the simulations and provided valuable feedback. Through a children's competition for the best quote from the first astronaut on Mars and the best design for a Moon mission flag, the interest of the young generation in Space exploration and human spaceflight was ignited. Overall, the massive public and media interest in the MOONWALK simulations is proof that Space exploration is still a fascinating topic and that earth analogue simulations in Europe are not only valuable to validate key technologies, but also to efficiently promote human space flight .
Project Context and Objectives:
Today, the exploration of Moon, Mars and other objects in Space, such as asteroids, is done exclusively with the help of robots and automated probes. NASA’s Curiosity mission is a recent example of a successful robotic exploration mission to Mars. With ExoMars, ESA has plans for a European robotic mission to Mars in 2020, and the Chinese sent a robot to the Moon in 2013.
However, the full potential of Space exploration can only be reached with human astronauts on-site, and thus both NASA and ESA are working on plans for human missions to Mars, Moon and asteroids in the years after 2020. In these missions, humans will be accompanied and supported by robots, and a close co-operation of humans and robots will be required.
The objective of the MOONWALK project was the development and testing of technologies for human-robot interaction during Extra Vehicular Activities (EVA) on planetary surfaces. The control of robots on a planetary surface is non-trivial because astronauts are limited in their movements by a bulky spacesuit and the effects of reduced gravity. MOONWALK developed new, practical methods for the control of robots by astronauts. These methods were tested in two earth-analogue simulations in a desert near Rio Tinto (Spain) and underwater off the coast of Marseille (France). In both cases, a planetary EVA was simulated under environmental conditions similar to those found on Mars (e.g. acidic soil chemistry, rugged terrain, dust contamination, communication delay) and the Moon (e.g. reduced gravity).
In both trials, an astronaut was made to wear a simulation spacesuit that imitated the properties of a real pressurized spacesuit, like weight, movement constraints, and visual constraints. The astronaut was provided with a small helper robot and newly developed interfaces and tools to enable robot control, communication with the local mission control (e.g. in the lander) and the mission control center on Earth, and on-site soil sampling and science.
Seven beneficiaries from seven EU member states participated in MOONWALK: The DFKI Robotics Innovation Center in Bremen (Project Coordinator), COMEX in France (technical coordination), EADS Airbus Defense Systems in the UK, LIQUIFER Systems Group in Austria, Space Application Services in Belgium, NTNU Centre for Interdisciplinary Research in Space in Norway, and the Instituto Nacional de Técnica Aeroespacial (INTA) in Spain. Intensive contact to external stakeholders, such as ESA and NASA, were established during the project.
The overall objectives of the MOONWALK project can be summarized as follows:
- Develop innovative technologies to support co-operative human-robot exploration of planetary surfaces (robot-assisted EVA);
- Use and further develop infrastructures, technical systems, and know-how needed for analogue simulations in Europe;
- Test and demonstrate the technologies developed in MOONWALK in two analogue simulations (in Rio Tinto and Sub-Sea Marseille) under simulated Mars and lunar conditions;
- Reach out to the (European and international) Space community and to the public to boost the interest in human space flight and in European analogue simulations.
The following is a list of the main MOONWALK objectives and their requirements, and how these objectives were addressed and met in the MOONWALK project:
O1 requirements: Human-robot and human-human cooperation in extreme environments with shared robot control between the Control Centre and the astronaut on-site. This includes wearable Human-Machine Interfaces (HMI) that work in extreme environments and under difficult operation conditions.
O1 achievement: Various human-machine-interfaces (wrist display, tablet, gesture control) and a control center were successfully implemented and tested in relevant environments (EVA simulations). Proof-of-concept that the interfaces work under difficult conditions was provided, albeit at a low TRL and without full integration of direct robot control through.
O2 requirements: Adaptation of an (existing) autonomous operating rover-type robot platform for human-controlled interaction inclusive required instrumentation (cameras, sensors, communication and navigation package).
O2 achievement: The YEMO helper robot is a new construction based on the DFKI ASGUARD system, fulfilling the requirements of MOONWALK (amphibious operation in difficult terrain). Sensors integrated in YEMO included cameras and a communication package. Because the focus of MOONWALK was on astronaut-robot interaction and control and the budget was limited, the final decisional autonomy of the robot was low and limited to a follow-me functionality.
O3 requirements: Design and setup of communication, mission planning and operations infrastructure, as well as tools (control centre) which can be adapted to various mission scenarios (such as Moon or Mars with variable communication delays).
O3 achievement: A mission control infrastructure was set up with an on-site control center simulating CapCom, or the local control center on the planetary surface, and a remote control center in Brussels, simulating mission control (MCC) on earth. The set-up was successfully tested in both analogue simulations. Variable communication delays between MCC and CapCom were implemented.
O4 requirements: Evaluate human performance in extreme environments (in function of gravity level variations, pressure and temperature) and the estimate the effect of those environments on robot-human cooperation. The goal was to establish lessons learned for the design of future exploration missions and associated technology, and to identify the factors and stressors which affect human performance and to develop mitigation strategies.
O4 achievement: Two earth analogue simulations were performed in a desert-like environment (Rio Tinto, Spain) and underwater off the coast of Marseille to simulate extreme environments as much as possible within the framework of the MOONWALK project. Human performance (of mock-up astronauts) was measured during the analogue simulations, both with technical means (bio-monitoring) and with empirical means (questionnaires). Lessons learned were extracted and formulated in deliverables D8.1 and D8.2. Although these results prove that robots can be valuable helpers for humans even in extreme environments, the overall scope of the analogue simulations possible in MOONWALK was too limited to support a scientifically sound empirical assessment to identify stressors and develop mitigation strategies.
O5 requirements: Improve the sustainability of life in extreme environments through protection garment and portable life support systems for astronauts. The devices tested in MOONWALK should include a bio-monitoring system integrated in the EVA space suit to monitor basic life signs of the astronaut during the EVA. Physiological correlations between crew activity, life support suit performance, crew health and subjective well-being of the mock-up astronauts should be established.
O5 achievement: A new EVA simulation space suit that can be used for both terrestrial and underwater simulations was designed and build in MOONWALK. For the simulations, a bio-monitoring system was integrated in the EVA suit. The collected data was used to monitor the health performance of the astronauts during the analogue simulations (see D8.1 and D8.2).
O6 requirements: Define new robot-supported search methodologies and strategies to detect extremophile life forms and bio-signatures in terrestrial analogues by integrating existing hardware in the mission scenarios.
O6 achievement: Devices for in-situ analysis of soil probes, a RAMAN spectroscope and the SOLID in-situ lab tool were customized to be carried by the YEMO helper robot. The tools and handling by the helper robot could prove the general functionality and usability of the concept. However, due to technical problems during the simulations, the full potential of the robot-mounted tools could not be shown.
O7 requirements: Develop manual and robotic sampling tools and field exploration procedures and test them in extreme environments (low temperature, microgravity) and for different application fields (geology and exobiology). Assess the effectiveness of the tools when used by astronauts in extreme environments.
O7 achievement: Several manual sampling tools and robot-mounted support tools were developed and implemented. A set of field exploration procedures was defined and used to test the tools in the analogue simulations. The usability of the tools by the astronauts was evaluated empirically. Unfortunately an active robotic sampling tool (e.g. manipulator arm) could not be implemented due to budget constraints, which limited the usability of the robot for sample collection. Also, a scientifically sound quantitative evaluation of the efficiency gains created by the manual tools could not be provided due to the limited scope of the analogue simulations.
Project Results:
The main results of MOONWALK were A.) technologies and tools, B.) new capacities related to European infrastructures for earth-analogue simulations, and C.) data on the feasibility of human-robot interaction on extraterrestrial planetary surfaces.
A.) With respect to technologies and tools, the main results were:
EVA simulation space suit: COMEX developed a concept and a design for a new EVA simulation and training suit. The design was influenced by previous work and experience of COMEX with the GANDOLFI EVA training suit. However, efforts were made to take also new designs of NASA suits into account: The new training suit was based on the same architecture and anthropometric capabilities of the NASA Z-series spacesuits prototypes and includes a rear entry system as the Russian ORLAN suit. In addition, the concept took into account the requirements for robot-assisted EVA identified in MOONWALK, the results of human factor analysis, and the operational requirements of the planned Earth Analogue Simulations.
The EVA training suit incorporates an exoskeleton made of hard and soft parts, and a diving gear for the astronaut diver for the underwater configuration. The exoskeleton design had three main drivers, to take into account ergonomics aspects of the Z-1 suit (NASA input to MOONWALK), to simulate constraints of movements as in a real pressurized space suit, and to put the astronaut/diver into neutral buoyancy (or reduced gravity). This function is only available in underwater configuration.
The exoskeleton is composed of hard shells made of composite materials that enable a significant weight gain compared to standard metal materials with a high mechanical resistance. Those parts are linked with neoprene parts.
The second major hard elements are the bearings allowing rotations of the different limbs, whom rings are made of POM and ball made of 316L steel (AISI standard).The bearings location is directly inspired by the Z1 series American spacesuit to allow rotations of shoulders, arms, thighs and pelvis.
These hard parts are linked through soft neoprene parts, tailored-made manufactured by Beuchat company, fixed on bearings.
To simulate the movement-constraints caused by a pressurized spacesuit the movement amplitudes of the training suit were limited by the design itself to comply with the Z1 spacesuit movement constraint specifications. In addition, a mechanical resistance was induced on the joint locations. The joint elements on the leg and arm assemblies (knee and elbow) restrain movements by creating a resistant torque. This makes the flexion of the members harder. The joint elements are torsion springs, which are sized to be the closest of the Z1 spacesuit movement resistances.
Astronaut subjects using the EVA suit for MOONWALK are limited to the following heights: minimal 160cm to maximal 195cm. To adjust arm lengths and leg lengths, adjustment tabs are located on both elbow and knee joints as shown below. To adjust the torso-pelvis distance, the length strap located on the torso can be changed according to the subject’s height.
The retractable visor is made of PMMA (ALTUGLAS®) and fixed on the helmet authorizing a rotation for allowing it to be used in down or up position. The field of vision has been optimized through the helmet design and the full face mask performance and can theoretically reach 60°.
The design of the gloves for the space suit was based on current NASA EVA gloves. The gloves were manufactured from a neoprene classic diver gloves recovered by the outer fabric of standard ski gloves. Home-made silicon parts were manufactured and stuck on to the gloves outer skin in order to make them look like NASA gloves.
Human-Machine Interface (HMI) and EVA Information System: The main features of the HMI were a chest display and a wrist display, both developed by Space Application Services. The chest display consisted of a tablet computer which is located on the torso of the space suit. The touch-screen of this device was modified to work underwater. The interface displayed on the chest display was set up to enable the use with heavy space gloves. The chest display was mainly used to provide information related to the mission procedures (recorded in standard NASA format) to the astronaut.
The wrist display was intended a.) to replace standard US EVA cuff checklist on the spacesuit and b.) to provide an alternative to the gesture control (see below) of the robot. This display used a small a screen to display an interface consisting of an array of simplified push buttons.
A Suit Computer Assembly (SCA) was integrated in the space suit to enable communication between the astronaut and the local mission control center. The SCA was a computer unit in a waterproof housing that does the necessary processing for the Human Machine Interfaces, but also for other potentially required sensors, including the Rover Gesture Control. In the underwater simulations, the SCA connected to a communication buoy via an umbilical. In the Rio Tinto simulations, an wifi connection was used to connect to CapCom.
A Mission Control Centre was set up at Space Applications in Brussels, Belgium. During the simulations, this control center simulated the MCC on earth. Consequently, a time-lag of up to serveral minutes between MCC and CapCom was implemented to test the effects of delayed response and local vs. remote decision making..
Robot control-by-gesture: In addition to “traditional” HMI solutions, MOONWALK investigated the feasibility of an innovative way to interact with robots. The approach developed by DFKI uses simple gestures made by the astronaut with his/her upper limbs. The gestures are recorded with IMUs (Inertial Measurement Units) embedded in the space suit, interpreted with a special software, and then translated to control commands for the robot. Using IMUs instead of optical sensor (cameras, 3-D cameras), which is a more common approach, has several advantages:
- The gesture recording is independent of variable visibility conditions. This is important for using the system in earth analogue simulations underwater, but will also play a significant role in a potential application in future Mars missions;
- The IMU’s are water-proof and can thus be used in underwater analogue simulations;
- The astronaut does not have to be in the line-of-sight of the robot, i.e. the astronaut can control the robot independent of his/her relative position to the robot and also record the pose of the astronaut in emergency situations (i.e. astronaut lying on the ground).
In MOONWALK, a simple set of gestures was implemented and tested, both in the lab and in both earth analogue simulations. The method proved to generally work very well and, after some initial training, was described by the astronauts to be easy to use and practical. However, improvements in the reliability and performance of the underlying software are still necessary. For the application of the gesture control in terrestrial spin-out applications (e.g. for the control of industrial robots), the enlargement of the gesture repertoire is planned.
Amphibious helper robot: DFKI used the general design of their well-tested and proven ASGUARD robot family as the basis for the MOONWALK robotic helper. This concept was modified to accommodate the requirement of a system that can operate both on land and underwater. The result, the amphibious YEMO rover, featured a water- and pressure-proof (up to a water depth of 10 m) body with four individually actuated star-wheels or regular wheels. The weight of the robot under terrestrial conditions was around 15 kg, but could be adjusted to simulate lunar conditions underwater. The size of the robot is approx. 1,5 m x 0,4 m. With the four star wheels, this small rover can climb steep slopes and handle difficult terrain.
The main sensor integrated in the rover was a watertight 360 degree camera. With this camera, a constant stream of 360 degree images can be created and analyzed. The set-up enables several analysts in the mission control center to analyze different parts (e.g. view angles) of the data in parallel.
A payload box developed by Liquifer Systems Group was mounted on the rover. It contained the manual tools and was designed to hold soil and rock samples. It also contained the RAMAN spectrometer, which was modified to fit on the robot and in the payload box.
Biomonitoring system: The underlying principle of the Biomonitoring system developed by EADS / Airbus was that the astronaut’s physiological signs can be monitored to improve the overall picture of their welfare. This is achieved through several systems: the monitoring of the user’s heart rate (through traditional beats per minute (BPM) recording) and the assessment of the physiological and mental stresses being experienced by subjects.
The system uses a machine learning approach with classifiers to determine, from real life data, the physical state of the astronaut. As this is a statistical approach, there is not a requirement to understand the underlying physiological nature of the readings you are gathering, but instead to recognize patterns and behaviors in the underlying data stream and how they relate to previously acquired data.
In order to build this system, appropriate statistical features had to be identified, which was a key focus for the research in MOONWALK. Prior research on the subject, and the experience of the research team, has focused on Heart Rate Variability (HRV) as the primary source for these features. HRV is the change in time between each heartbeat, and is affected by physiological and mental stress.
Many of the candidate features are focused on the frequency domain, understanding the various components that make up the overall ‘signal’ of the Heart Rate Variability. These features can be fused together, such as ratios between high frequency and low frequency components.
After identification of appropriate features, the classification framework was ‘trained’ with test data acquired from various subjects in different conditions of physiological and mental stress, alongside a ‘ground truth’ to represent an indication of the stress levels experienced.
The current prototype system classifies the user’s stress into three broad categories: Red, Amber, Green. This ‘traffic light’ system acts as a general indicator, ‘Green’ being a relaxed, non-stressed state, ‘Amber’ indicating medium physical and mental stresses, and ‘Red’ meaning a high level of physical and cognitive loading.
Tools for manual, robot-assisted sampling: The robot operational scenarios developed in MOONWALK comprise a range of astronaut-robot interactions of various degrees of complexity. Liquifer Systems Group developed a number of manually operated tools to support the astronaut. In particular, these were an Astronaut Rescue Tool (ART), an Astronaut Tether Control (ATC), a Pantograph Sampling Tool (PST), a Foldable pick-up Claw (FPC), and a Manual Tool Rack (MTR). All the manual tools aim at safe operation through a single astronaut in cooperation with a robotic rover.
The Astronaut Rescue Tool (ART) for astronaut assistance should he/she fall over is a manual tool which is positioned in a separate compartment of the Payload Box (PB) installed on the robot. The collapsible tool is single-handed accessible to the fallen astronaut and can be deployed single-handed by flicking the arm. It is developed from an off-the-shelf walking stick with adapted handle and anti-slip elements along the stick. The Astronaut Tether Control (ATC) is the manual control for the rover scouting a steep slope. ATC is developed from an off-the-shelf retractable dog leash with adapted control buttons and modified handle to fit the astronaut glove. The end of the tether is modified for easy attachment to the rear of the rover.
The Pantograph Sampling Tool (PST) is a retractable sampling tool which can be used single handed to collect samples and feed them into sampling bags. The design is being developed from a lazy tong concept which allows elongation and contraction of the tool arm. It is equipped with a scoop gripper. The Foldable pick-up Claw (FPC) for sample collection is based on an off-the-shelf foldable system. Both collapsible tools can be transported in the payload box of the robotic rover or can be attached to the astronaut suit when feasible. The Manual Tool Rack (MTR) replaces the manual tool cart which was foreseen for (sampling) tool transport in earlier scenario versions. The rack simulates a rack for bigger existing tools (supplied by COMEX), sampling bags and construction material similar to the Lunar Roving Vehicle (LRV) rack. It can also serve as support to secure the Astronaut-rover tether operations with a winch system.
Using rapid prototyping and 3-D printing, workable prototypes of the manual tools were developed and tested in the earth analogue simulations in Rio Tinto and Marseille.
B.) With respect to the earth analogue simulations, the main results were:
Rio-Tinto Simulation: The technical equipment, methods and mission procedures developed in MOONWALK were verified in a Mars analogue simulation in Rio Tinto (Spain) between April 16th an 30th 2016. A planetary exploration mission was simulated in which a team of an astronaut and a robot executed different tasks and scenarios required to achieve the mission objectives: i) Find a safe and secure site for human settlement; ii) Finding resources: minerals, material for contractions (e.g. gravel, rocks); iii) Astrobiological research by searching for signs of live. The performance of the astronaut-robot team was compared qualitatively to the performance of an astronaut-astronaut team doing the same EVA tasks.
The whole campaign lasted two weeks. The first week was devoted to setting up, testing and training all the equipment and procedures. The planetary exploration mission simulation was performed during the second week. The main elements for the simulations were: a prototype of a Martian habitat (SHEE habitat), the Gandolfi II astronaut training suit, a small helper robot (YEMO) gesture-controlled by the astronaut and equipped with different tools developed during MOONWALK, and scientific instrumentation for astrobiological investigations.
An average of 25 persons from the different partners of the MOONWALK consortium participated directly in the different aspects of the campaign. Among them, a medical team for medical check of the simulation astronauts, as well as the participants in the external experiments.
The main achievements of the campaign were: i) Multiple EVA simulations involving only astronauts or hybrid astronaut-robot teams were carried out to test and execute the tools, procedures and scenarios developed during the MOONWALK project, including gesture-controlled astronaut-robot scouting, sampling, or exploring inaccessible sites by any of them separately; ii) A realistic Mars exploration mission simulation was performed with the main objectives achieved: the landing site was explored and mapped, different resources were identified, and microbial markers (evidences of life) were detected in the collected samples; iii) extensive outreach and dissemination of the campaign both in the field and through press, TV, radio, both national and international (See D9.6).
Marseille simulation: The simulation campaign in Marseille has been set up accordingly to one of the MOONWALK project’s task definitions: “The objective is to perform a simulated Moon mission at an analogue site offshore Marseilles. The objective of these tests will be to validate the human-robot interaction architecture and to simulate various activities that future astronauts might perform on the Moon (cave exploration by robot, crater descent, installation of instruments and mock-up habitation structures. “
Nine potential lunar analogue sites were identified in the area of Marseilles, in the Calanques National Park. Situated at a reasonable water depth, they were selected mainly for their geomorphological similarity with some interesting spots on the lunar surface. Among those sites Port de Pomègues was selected for the MOONWALK lunar surface EVA simulations. The reasons for that choice were:
1. The site hosts several geological features of interest, namely a crater-like structure (which in reality is a sinkhole) and steep edges on the borders of the site.
2. Its exposition to wind is low (which is quite important for the Marseilles area where strong winds can quickly block such tests for several days) and the operations vessel MINIBEX can navigate in the zone.
3. A preliminary survey on the strength of communication signals showed that this site is best covered amongst the nine sites identified.
The underwater analogue simulations in Marseille were conducted between May 30th 2016 and June 10th 2016. Because of the complexity of the simulation architecture and of diversity of equipment involved and managed by the different partners, preliminary EVA pool tests were necessary to train the divers who were selected to be astronauts and experience for subsea operations (particularly for security procedures rehearsal). Thus the week from the 30th of May to the 3rd of June was dedicated to pool trials in the COMEX’ facilities. The second week from the 6th of June to the 10th of June was devoted to the subsea analogue simulation to Moon with the purpose of testing the Astronaut-Robot and Astronaut-Astronaut procedures.
Human factor evaluation: In MOONWALK, analogue simulations we evaluated the human performance and the effect of robot-human cooperation in very different environments and in interaction with different simulation components, i.e. astronaut suit, astronaut tools, human machine interfaces (HMI). The simulations generated several types of data; video and audio recordings, GPS tracking, biomonitoring, written logs and questionnaires and interviews. However, due to the considerable amount of data, technical challenges and time limitations the human factor evaluation are limited to data and information from written logs (CAPCOM), questionnaires and interviews.
The findings of the trials indicate that the different sites lead to differences in how the astronauts evaluate the system components and their experience of the simulations. However, there is little indication that the interaction with the rover/robot resulted in different responses from the simulation astronauts compared to interaction with another astronaut, i.e. the robot was seen by the astronauts as a valuable and useful helper.
Generally, simulations in analogue sites have shown to represent great tools to learn about the complex socio-technical system that a human exploration mission will represent. However, when comparing terrestrial (Rio Tinto) versus underwater (Marseille) simulations, the main conclusions are that
- terrestrial simulations are a very useful mean to work on the technical aspects of a complex simulation architecture. It allows engineers easily to stand next to the simulation-astronauts, try out tools, operations and procedures and change those based on the feedback from the field. Test can be quickly performed and adapted. Terrestrial simulations are therefore very useful to improve technologies and do engineering on EVA equipment.
- subsea simulations are less useful for such engineering considerations (or “debugging”). Once all the materials is on the ship or underwater all has to work. Delays and small errors lead to large delays in the operations since various security issues have to be respected. However, from feedback from the subjects that served in both simulations a clear feedback was that the subsea simulation is much more realistic in terms of technical difficulty and mental stress. Subsea simulations are therefore very useful for training purpose of astronauts.
Conclusions:
- All the MOONWALK elements were fully deployed and have been fully operating: the robot YEMO, the training EVA suit Gandolfi II, the tools for collecting and sampling, the biomonitoring system, and the communication system.
- This first attempt of a European subsea analogue simulation with a robot went beyond the proof of concept, and had shown the relevance of team/robot cooperation through the realistic procedures scenarios. Invited visitors from ESA confirmed this interest in such simulations in Europe in the future.
- The outreach was extremely successful with a world-wide coverage of still photography and video material.
Potential Impact:
Impact: MOONWALK developed and evaluated new technologies for human-robot interaction, built know-how and equipment in support of European capabilities for earth analogue simulations, and raised the awareness for human spaceflight and the benefits and opportunities of earth analogue simulations with European and international stakeholders as well as with the general public.
On a political level, MOONWALK helped to create an understanding for the necessity to build and use European infrastructures for earth analogue simulations to test space technologies and prepare for future missions with a justifiable effort. The project also helped to raise the public awareness of space exploration and to re-ignite the fascination in human space flight, in particular in the young generation.
Overall, MOONWALK was a showcase project which not only inspired people to make the next step into human robotic space exploration but actually already connected to one of the possible end costumers, ESA, for the inclusion of MOONWALK products into their studies and programmes.
Exploitation: On a technical level, the project developed and tested a number of technologies and equipment that will be available for future earth analogue simulations and be scientifically and commercially exploited by the project partners for space-related and terrestrial applications.
The general MOONWALK exploitation foresees that all partners are allowed and encouraged to exploit, both commercially and non-commercially, any technical results, know-how, or facilities they developed during the MOONWALK project as they wish. Legally, any exploitation activity has to conform with the rules for the distribution of intellectual property rights (IPR) as defined in the Consortium Agreement (CA). These relevant paragraphs of the CA are listed in Appendix A.
Overall, the CA adopts a policy that gives IPR on knowledge and/or solutions (technologies) to those partners that developed the knowledge / solutions within the project. Where several partners worked on the same exploitable results together during the project, specific conditions for sharing these results (licenses) have to be negotiated between the partners.
In terms of exploitation of project results by individual consortium members, there are generally two different types of organizations that use different general exploitation strategies:
• University and Research Partners have a focus on the non-commercial exploitation of project results. The primary exploitation channels for these partners are to
o use project results in on-going national and European R&D projects;
o leverage project results, contacts and expertise gained in the project to apply for new R&D projects;
o to include project results in the graduate (and undergraduate) teaching;
o to use scientific project results for on-going or new PhD projects;
o to market high-TRL results through spin-off companies (if feasible as joint-ventures with other project partners).
• Industrial Partners are mainly interested in the commercial exploitation of project results. Their main channels for exploitation are related to a
o further evaluation and progressive prototyping of the promising project results;
o integration of specific technologies developed in the project in existing products;
o the “spinning off” of new products.
Almost all of the technical outcomes described in the previous section can be applied in terrestrial applications. An example is the control-by-gesture, a method that is already exploited by DFKI in support of applications related to Digital Industry (i.e. robot-control in an industrial environment). Another example is the bio-monitoring developed by EADS / Airbus which is used for operators in extreme environments.
Other outcomes will be exploited in future space-related projects and applications. An example is the simulation space-suit Gandolfi II which will be offered by COMEX to future earth analogue simulations as part of R&D projects or campaigns preparing for future space missions.
Dissemination: The outreach and dissemination goals of MOONWALK were to establish a high visibility of the project not only within the scientific research community, but also on a much broader scale that includes the general public, the (space) industry and the stakeholders (including European and international space agencies).
The idea was to educate and enlighten the target audience about the possibilities of working with machinery in extreme environments, highlight applications of space technologies on earth and attract future research, academic and industrial partners for further development (including commercialization) of the concepts and prototypes created through MOONWALK.
As outlined in Final Dissemination Report (D9.6) the consortium partners under the lead of Liquifer Systems Group have used a variety of means and tools to achieve these goals. This included a well-designed and maintained web site, four public newsletters, flyers and fact-sheets, and an extensive project video summarizing the approach and results of the analogue simulations.
The project was presented with scientific conference papers at various international conferences, including IAC 2014 and 2015, IAASS 2014, ASTech 2014, Astrobiology Science Conference 2015, AIAA Space 2015, ESREL 2015, International Symposium on the Moon 2020-2030. In addition, papers in Astrobiological and Space-related journals were submitted. There was significant media coverage of the project and the analogue simulations, with more than 50 press articles and TV reports (see D9.6 Final Dissemination Report for details).
In addition, the project managed to create significant synergies with other EU-FP 7 projects. An example was the SHEE (Self-deployable Habitat for Extreme Environments) project (http://www.shee.eu/main). The planetary habitat test-bed for terrestrial analogue simulations was used at the MOONWALK simulations in Rio Tinto for an astrobiology laboratory and the Local Mission Control. Specifically for the MOONWALK project a suitport was built to the SHEE habitat to simulate ingress and egress procedures for the first time in Europe.
Dissemination of MOONWALK towards the European and international space agencies was high on the agenda of the MOONWALK team. MOONWALK connected and cooperated with NASA for the exchange of ideas and has enabled a feedback loop that has proved useful to both. There has also been a strong tie to ESA with regard to Moonwalk components, for example the suit, the tools, the communication architecture amongst other parts. There are two study projects with ESA, one was called “LUNA Analogues for Preparing Robotic and Human Exploration on the Moon – Needs and Concepts” [RD12], the other one is called MOONDIVE (ongoing). Both are dedicated to artificial analogues in preparation of future lunar exploration as a described by the ESA DG Jan Wörner in his Moon Village concept. In both concepts MOONWALK hardware is integrated and become key components of the ESA concepts. MOONDIVE is a study about scenarios for the Neutral Buoyancy Facility at ESA-EAC with the Gandolfi 2 suit to train astronauts for future lunar exploration. LUNA is about a whole artificial analogue to be established at ESA-EAC to support astronauts in training for future lunar or Martian surface missions. Furthermore, MOONWALK was introduced at the ESA Isolation Steering Committee meetings as a facility ready to be used for future ESA isolation studies.
MOONWALK supported a number of strategic activities, including the EC Analogue Workshop held in Brussels on October 12th 2016.
List of Websites:
MOONWALK Website: http://www.projectmoonwalk.net/
Coordinator contact: Dr. Thomas Vögele, DFKI Robotics Innovation Center, Robert-Hooke Str. 1, 28359 Bremen (Germany), thomas.voegele@dfki.de
