LOLA, humanoid robot , Technische Universitat Munchen, Munich Germany

[Airicist]

wevolver.com/wevolver.staff/lola (https://www.wevolver.com/wevolver.staff/lola)

mw.tum.de/en/am/research/current-projects/robotics/humanoid-robot-lola (https://www.mw.tum.de/en/am/research/current-projects/robotics/humanoid-robot-lola)

youtube.com/AppliedMechanicsTUM (https://www.youtube.com/AppliedMechanicsTUM)

Playlist "Humanoid Robot LOLA (https://www.youtube.com/playlist?list=PLVAvoOVYkpkMFNgz8cercTVvVryz1eh0a)"

[Airicist]

https://youtu.be/P4Y41Ago3cg

Humanoid Robot Lola at Hannover Fair 2010

Uploaded on Jan 14, 2011


At the Hannover Messe 2010 the humanoid robot Lola was presented to the public. Lightweight design, modern drive technology, a powerful sensor system, microelectronics and a dynamic walking control system enable stable biped locomotion. Two high-resolution cameras and an advanced image processing system give Lola the ability to navigate autonomously in an
unknown environment. The robot has been developed at the Institute of Applied Mechanics (AM) at the Technische Universitat Munchen.

At the Hannover Messe 2010, Lola is equipped with a vision system developed by the Institute for Autonomous Systems Technology (TAS) at the University of the Federal Armed Forces.

The combination of vision system, mechatronics and control system gives the robot a high degree of autonomy.

Lola was designed as an improved version of the humanoid robot Johnnie which was also developed at AM and presented to the public in 2001. Lola is 180 cm tall and weighs approximately 60 kg. As with Johnnie, Lola's proportions are derived from human anatomy.

For the robot's development, a wide range of new ideas and technologies was realized. Additional actuated joints in toes, pelvis and arms are key to more stable, versatile and human-like walking. A hierarchical control system
gives Lola the ability to walk in arbitrary directions like e.g. walking sideways to pass narrow points.

The robot visually perceives its environment through two high-resolution video cameras each with 5 megapixels. The incoming image is partitioned into areas of dierent importance. The less interesting areas are only examined roughly while special attention is drawn to the crucial areas. That way, expensive calculations are only carried out in areas in which new information is needed. This new method of attention based object recognition allows for a detailed analysis of the observed objects while at the same time reducing computing time.

At the Hannover Messe 2010, Lola navigates in an environment containing objects of every day use. The object layout can be varied by spectators and is therefore not known to the robot beforehand. In this scenario Lola will autonomously complete certain predened tasks.

[Airicist]

https://youtu.be/VZyrpoK8X7o

First Movements of LOLA, our Humanoid Robot

Published on Mar 25, 2015


Initial experiments with Lola, hanging from the safety harness. Special focus on the joint actuators. 2009.

[Airicist]

https://youtu.be/oSTnL3Zcxzg

Humanoid Robot LOLA avoids falling down

Published on Mar 25, 2015


An Estimation Model for Footstep Modifications of Biped Robots. IROS 2014. Our Humanoid Robot "Lola" reacts to unknown external disturbances and modifies its footstep positions accordingly.

[Airicist]

https://youtu.be/wC5__s7LhOE

Humanoid Robot LOLA stepping on unmodelled obstacles

Published on Mar 25, 2015


Videos of our Humanoid Robot "Lola" stepping on unknown and unexpected obstacles, with no vision information. The gait phase is switched using the force/torque sensors on the ankles. 2011.

[Airicist]

https://youtu.be/CwdQjO5_OhI

Development of Humanoid Robot LOLA

Published on Mar 25, 2015


Video showing the different aspects of the development of our latest Humanoid Robot, together with some of its first steps and accomplishments. 2012.

[Airicist]

https://youtu.be/ogZ9oBdCNU0

Humanoid Robot LOLA detects obstacles with a Kinect-like sensor and finds a way around them

Published on Jul 21, 2015


Our Humanoid Robot "Lola" uses an on-board Asus Xtion PRO LIVE rgbd sensor for detecting and approximating obstacles, and uses an A* algorithm to find a way around them in real-time. 2015.

[Airicist]

https://youtu.be/6PLN6B4vSHM

Our Humanoid Robot LOLA approximates moving objects with a Kinect in complex environments

Published on Jul 21, 2015


Using a Kinect-like sensor, our Humanoid Robot "Lola" is able to detect, approximate and track moving objects for avoiding them. Our motion planning system uses simple geometries to check for collision-free paths.

[Airicist]

https://youtu.be/46YIWkYWuoY

Humanoid Robot LOLA walks while being punched and avoids falling down

Published on Jul 21, 2015


A non-linear prediction model for footstep modifications. Our Humanoid Robot "Lola" reacts to unknown external disturbances and modifies its footstep positions accordingly.

[Airicist]

https://youtu.be/RjqAh3Blxng

Robust and versatile bipedal walking in unknown environments

Published on Jan 11, 2017


In this video, we show the combination of our vision, motion planning and stability modules on our humanoid robot Lola. This framework achieves autonomous biped navigation in unknown environments from simple walking commands.
Obstacles are modelled as one or more capsules. Then, the footsteps are planned by taking into account the obstacles' location and performing real-time 3D collision avoidance. While the robot is walking, its sensors watch out for un-modelled disturbances. To avoid falling down, footsteps locations are modified on-the-fly inside of collision-free regions.

[Airicist]

https://youtu.be/QiI7C_4lU7E

Film of LOLA (2010)

Published on Jan 12, 2017


Fast, robust and autonomous walking: demonstrating the initial capabilities of our Humanoid Robot "Lola".

[Airicist]

https://youtu.be/rKsx8HKvBkg

Humanoid Robot LOLA walks over an unexpected platform

Published on Jan 14, 2017


Vision-Based Autonomous Walking over Platforms or Stairs.
- Our vision system uses a kinect-like sensor to model the environment: surfaces (including the ground) are modelled as polygons and obstacles as SSV's (swept-sphere volumes).
- Our motion planning system uses a modified A* search algorithm to find footsteps locations on walkable surfaces and each step's parameters are optimized to account for kinematic capabilities.
Everything is running in real-time

[Airicist]

https://youtu.be/g6UACMHgt20

DFG project final presentation: versatile and robust walking in uneven terrain

Published on Jan 18, 2017


After the completion of the DFG Project "Flexibles und robustes Gehen in unebenem Gelande", the LOLA-Team made a small workshop, with live demo included:
- fast walking
- joystick navigation
- vision-based, real-time navigation in presence of obstacles and unknown disturbances
- vision-based, real-time navigation over platform
- stability under external disturbances

[Airicist]

https://youtu.be/-VvxzFg9ATU

Humanoid Robot LOLA navigates towards a goal among moving obstacles

Published on Feb 14, 2017


In this video, we show an extended path planner for our humanoid robot LOLA.
First, 2D safe paths are generated, avoiding large obstacles which cannot be traversed. Then, the A* planner finds a safe way along those paths, taking care of stepping over obstacles.
The system is able to react to unknown dynamic environments, like a person walking in front of the robot.

[Airicist]

https://youtu.be/CwdQjO5_OhI

Development of humanoid robot LOLA

Published on Mar 25, 2015


Video showing the different aspects of the development of our latest Humanoid Robot, together with some of its first steps and accomplishments. 2012.

[Airicist]

https://youtu.be/VceqNJucPiw

Vision system of humanoid robot LOLA: Platforms and dynamic obstacles

Published on Oct 23, 2017


In this video, we test the capabilities of our Vision System and Motion Planning System, by letting the robot walk in previously unknown dynamic environments. The vision system approximates "walkable" surfaces as polygons and obstacles as capsules or spheres and the motion planning finds safe paths in real-time.

[Airicist]

https://youtu.be/EeDR1UNDpIY

Robot LOLA autonomous walking compilation + HoloLens Mixed Reality App

Published on Oct 29, 2017


In this video we show a compilation of our research for the last 4 years on autonomous navigation of bipedal robots.

It is part of the DFG-founded project "Versatile and Robust Walking in Uneven Terrain" (German Research Foundation) and includes development in environment perception and modeling, motion planning and stability control.

Additionally, we give a sneak peak of our HoloLens Mixed Reality App, "HoLola" which will be published soon!

[Airicist]

https://youtu.be/R0S0SYyWRkE

Development of humanoid robot LOLA

Published on Dec 7, 2018


A performance enhanced humanoid that features a modular, multisensory joint design with brushless motors.

The electronics architecture is designed as an intelligent sensor-actuator network with a central controller. The new decentral components increase the system’s performance from a technological point of view. Additional DoFs are introduced to allow for more flexible and natural motions. The trajectory generation and control system is currently being developed, aiming for faster, more flexible and more robust walking patterns.

[Airicist]

https://youtu.be/ifuD-ETTi_I

A force control scheme for biped robots to walk over uneven terrain including partial footholds

Published on Mar 21, 2019


This video shows several experiments of walking over unknown, uneven terrain with the biped robot LOLA. Three different methods for ground-force control are experimentally validated in early, late and partial contact situations. The methods are combined to a force-control scheme to allow for the traversal of highly uneven terrain without vision-based information.


https://youtu.be/pmtKv8VEItY

The humanoid robot LOLA walks over uneven terrain without using visual information

Published on Mar 21, 2019


This video shows how the humanoid robot LOLA walks over uneven terrain without using any vision-based information. This reactive robustness to unknown terrain is achieved by a ground-force control scheme that handles ground-height variations and unplanned partial footholds with the environment.

[Airicist]

https://youtu.be/cNkQT2SUegE

The humanoid robot Lola walks outside

Published on May 24, 2019


This video shows our humanoid robot Lola walking on asphalt, grass, and cobblestone. All motions were generated online and on board of the robot. The cable is needed for power supply as the robot doesn't have a battery yet. Lola is safeguarded by a safety harness - this rope is loose during walking.

For this video, Lola was either operated by a human via a wireless controller or commanded simple objectives, e.g. "walk straight". The motion planner assumes a flat ground and the different ground properties are previously unknown to the robot's feedback control methods.

[Airicist]

https://youtu.be/piQm_oTYXIc

Smooth Real-Time Walking-Pattern Generation for Humanoid Robot LOLA

Published on Jul 3, 2019


This video demonstrates our new approach for planning smooth center-of-mass trajectories for biped walking robots. The method is based on quintic spline interpolation and collocation and generates dynamically and kinematically feasible motions in real-time.

Method/Testscenario:
Our humanoid robot Lola steps up and down a platform of 12.5cm height. The planned center-of-mass motion respects the dynamics and kinematic limits of the robot using simplified models. The complete motion lasts more than 17 seconds and is planned in less than 9 milliseconds (CPU only, single-core).

Simulation:
Custom multi-body simulation
Visualization with Blender (custom interface using Blenders Python API)

Experiments:
Planning and control of the robot runs in real-time (onboard). External communication only triggers a signal to start and stop walking. The vision system is not active, thus foothold sequence is predefined.

[Airicist]

https://youtu.be/mpDqMFppT68

Humanoid robot LOLA v1.1 - Hardware upgrade for multi-contact locomotion

Ocr 26, 2020


In this video we present recent efforts to make our humanoid robot LOLA ready for multi-contact locomotion, i.e. additional hand-environment support for extra stabilization during walking. We focus on the hardware changes of the upper body (the pelvis and the legs remain unchanged).

Comparison of degrees of freedom (DoFs):
The new upper body features two additional joints, alias "arm rotation", which significantly increase the reachable taskspace. This is essential for supporting against walls in the lateral proximity of the robot.

Assembly:
We explain the composition of the new torso and arm design. The torso contains core components, like the inertial measurement unit (IMU), the two onboard PCs (control and vision), bus couplers and power bridges. Note that the cabling is not modeled and thus not visible in the renderings.

Finite Element Analysis:
A finite element analysis of two exemplary structural parts is shown. These parts have shown to be critical and were developed in an iterative process. The whole robot is designed for minimal weight.

Taskspace Analysis:
The reachable taskspace of the new arm design is compared to the previous topology. We define the torso segment as fixed "base" and the center of the hand as tool center point (TCP). The volume describes the reachable region while the surface is colored according to certain metrics. These metrics indicate "how well" a certain point in 3D space can be reached by the TCP. For this analysis the taskspace is defined as the cartesian position of the TCP relative to the base.

This work is supported by the German Research Foundation (DFG, project number 407378162).

[Airicist]

https://youtu.be/DzuFXPc2U5c

Humanoid Robot LOLA - Robustness to unexpected ground height changes

Feb 27, 2021


In this video, LOLA reacts to undetected ground height changes, including a drop and leg-in-hole experiment. Further tests show the robustness to vertical disturbances using a seesaw. The robot is technically blind, not using any camera-based or prior information on the terrain.

Technical Details:
Walking speed: 0.5 m/s
Height of the board: 6.7cm
Ground height change leg-in-hole: 6cm
Negative ground height change platform: 9cm