Specific task system

Maintenance task

The maintenance task is mainly performed by the robotic arm which can carry out the operation requested using the gripper end effector. Guided by the com-puter vision system, the rover is able to recognize the points of interest where the operations will be performed, recognize the shape and color of the objects on the panel and place the gripper thanks to the inverse kinematics calculations. Once the end effector is placed over the required point, pre-recorded tool paths will perform the operation necessary. Proximity sensors around the chassis are fundamental for positioning the rover at the right distance from the panel and the proximity sensor placed on the gripper avoids the risk of damaging the electrical panel. Volt-age measurements on the electrical socket will be performed by an additional tool picked up by the gripper that contains the electronic circuit which performs the measurement and sends the data to the main board by I2C bus. The computer

4 – T0-R0: engineering model of an astronaut assistance rover for the European Rover Challenge 2018

vision system is trained on a real model of a similar panel, built to simulate the task in our laboratory.

Figure 4.24: Gripper with voltage measurement tool

Scientific task

The scientific task demands the most complex operations of the ERC 2018 com-petition with strict constraints deriving both from the comcom-petition requirements and from our assumptions. Initial research showed how many variables and un-predictable situations the rover may encounter and deal with. Even if the quality of the surface layer of the terrain matches expectations, the nature of the mate-rial present below the surface is unpredictable, resulting in the need for a versatile drilling system, able to work in a wide range of conditions. The system is able to drill in both soft and hard materials, even though the strongest material likely to be encountered in the competition is gypsum (Mohs scale of mineral hardness 2 ).

The tests for the drilling system included a wide range of soil types, including the specific soil made available by ALTEC facilities at the Mars Terrain Demonstrator.

The scoop end-effector with an integrable core drill and a collecting system were designed to perform this task. This solution is compatible with the choice of using interchangeable end-effectors and additional tools to perform the tasks.

The scooping system can apply enough shear force to cut and collect superficial soil inside the scoop. The inverse kinematic control, addressed by the recognizing algorithm of the computer vision, drives the scoop to the exact point marked on the ground. Once scooped, the payload will be carried inside the collecting boxes by the robotic arm and unloaded. In the meantime the arm camera will take pictures

of the area. The collecting boxes are provided with a group of sensors that can measure weight, temperature and humidity of the sample. Volume measurements will be performed thanks to a reference grid printed inside the box. Storing boxes have an automated lid that seals the samples.

The same routine will be followed to excavate the trench and the arm camera will perform measurements and acquire photographic documentation. This subsystem is able to perform the task in accordance with the rules requirements.

In case of deep sample collecting the drilling is performed by a core drill tool grabbed by the scoop and stored inside the sample container area.

This solution is not exactly in line with the rules requirements but has been chosen for consistency with our research on interchangeable tools and end effectors.

This core drill is operated directly by the arm and does not require any other actuator or support structure except the robotic arm itself and its own motors.

Due to the limited power of the wrist actuators and to the risk of causing ex-cessive vibration and stresses on the arm structure, the drill cannot reach high rotational speeds.

Since at the moment this solution doesn’t completely comply with the rules requirements correctly in uncoupling the stresses caused by the drilling and in the maximum drill depth , we assume the need to develop a new flexible and light weight drilling solution. In comparison with more traditional drilling solutions placed on the rover chassis, the advantage of the robotic arm is that it provides great reaching capabilities and permits drilling in sites located on slopes, vertical walls or craters as well as on the ground. There is also the possibility of carrying more than one drill, allowing sampling on a wide array of soils. The core drill body functions as a sample container and allows samples to be stored and insulated from the environment. The sample can then be extracted from the driller body and analyzed in a laboratory.

Although we are aware that our solutions do not fully satisfy all the requested performances at this stage, our work embodies the results of the intensive research the team has undertaken and we confirm our intention to develop efficient and innovative solutions in the near future.

Collection task

The caching task is performed with the gripper end effector mounted on the arm wrist. The computer vision uses OpenCV in order to recognize the cache shape and colour from the navigation camera image and registers the location on the map.

The path finding algorithm drives the rover to the cache location and then the arm picks up the cache guided by the inverse kinematics. A specific storage zone on the chassis is designed to protect the caches from damage during transport and re-entry of the samples. Once the unloading site is reached, the gripper grabs the handle of the container, unlocking the retaining mechanism, and leaves the container on the

4 – T0-R0: engineering model of an astronaut assistance rover for the European Rover Challenge 2018

Figure 4.25: Scoop and core drill joining

required point on the ground.

Navigation task

During the navigation task the robotic arm will be removed in order to lighten the rover and reduce stresses on the chassis when the rover is driving on bumpy terrain and steep slopes. In this task the capabilities of the mobility system are exploited in full. The mobility system is designed to overcome unexpected obstacles and significant slopes. The absence of motors on the two central wheels are not a limit since the traction of the motorized wheels is constantly monitored by the control software of our VESC drivers board.

Autonomously driven by the computer vision system the rover reaches the points received by coordinates on the map using the SLAM algorithm.