Differences

This shows you the differences between two versions of the page.

Link to this comparison view

Both sides previous revision Previous revision
Next revision
Previous revision
about [2019/06/03 16:32]
tgeijten
about [2019/06/12 12:09] (current)
tgeijten [Who is SCONE for?]
Line 1: Line 1:
 ====== What is SCONE? ====== ====== What is SCONE? ======
-SCONE is open source software ​that aids in performing ​**predictive simulations** of biological motion+SCONE is open source software ​for performing ​[[doc:​predictive|predictive simulations]] of biological motion.
- +
-Predictive simulations do not rely on recorded motion data to estimate muscle force or joint torque. Instead, they compute motion trajectories that perform a given task optimally, according to high-level objectives such as stability, energy efficiency and pain avoidance.+
  
 Predictive simulations enable powerful new applications for musculoskeletal models, such as predicting the outcome of treatment, and optimizing the efficiency and efficacy of assistive devices. More fundamentally,​ it enables researchers to pose true //​what-if?//​ questions, allowing them to investigate the effects of individual model and control parameters on the motion as a whole. Predictive simulations enable powerful new applications for musculoskeletal models, such as predicting the outcome of treatment, and optimizing the efficiency and efficacy of assistive devices. More fundamentally,​ it enables researchers to pose true //​what-if?//​ questions, allowing them to investigate the effects of individual model and control parameters on the motion as a whole.
  
 ====== How does SCONE work? ====== ====== How does SCONE work? ======
-In SCONE, everything needed to perform a predictive simulation is bundled in what is called a **scenario**. Each SCONE scenario consists of the following components ​(click to expand):+In SCONE, everything needed to perform a predictive simulation is bundled in what is called a [[doc:scenario|Scenario]]. Each SCONE scenario consists of the following components:
  
-<​accordion collapsed="​true">​ +  * A [[doc:​model|Model]] of the human, animal or robot you wish to simulate. 
-<panel title="​Model">​ +  * A [[doc:controller|Controller]] that generates input for the model actuators
-The central part of a predictive simulation is the (musculoskeletal) model of the entity ​you wish to simulate. ​SCONE is designed to work with [[https://​opensim.stanford.edu|OpenSim]] for modeling and simulation. Any OpenSim model is supported -- both 2D and 3D -- as long as the model contains **controllable ​actuators**, such as muscle-tendon units or torque-based actuators. For simulating contact reaction forces, a model also requires **collision shapes** and **contact models**Finally, models typically contain ​**limit forces** to simulate ​the forces generated by tendons and bony structures when joint reaches ​certain limit.+  ​An [[doc:​objective|Objective]] that describes the goal task for which you wish to optimizethrough ​weighted combination of [[doc:​objective|Measures]]. 
 +  ​An [[doc:​optimizer|Optimizer]] that optimizes ​the free [[doc:​parameters|Parameters]] in scenario for specific objective.
  
-Always try to keep your model as simple as possible! Each actuator you add to the model requires additional control parameters that need to be optimized. Each element you add to a model makes the simulation run slower. +To learn more about SCONE scenarios, ​be sure check out the [[tutorials:​introduction]] tutorial.
-</​panel>​+
  
-<panel title="​Controller">​ +====== Who is SCONE for? ====== 
-Controllers compute the input values ​of the model actuators. They exist in two flavors+SCONE is designed to cater to a wide range of potential users, including
-  * Feed-forward (or open-loop) controllerswhich generate fixed patterns based on a parameterized function +  * Clinical researchers with limited technical skillswho wish to perform //what-if// scenarios using [[tutorials:​start|existing SCONE Scenarios]]. 
-  * Feedback (or closed-loop) controllerswhich generate actuator inputs based on sensor information+  * Biomechanics / neuromechanics researchers studying neuromuscular control. 
 +  * Robotics researchers interested in optimized control strategiesor the interaction between humans and assistive devices.
  
-SCONE uses a modular system in which controllers can combined to perform complex behaviors. +For more information,​ see the [[faq]]or post question ​on the [[https://simtk.org/plugins/phpBB/indexPhpbb.php?group_id=1180&​pluginname=phpBB|SCONE User Forum]].
-</​panel>​ +
- +
-<panel title="​Objective">​ +
-The goal of a predictive simulation is to find the motion pattern for which a specific task is performed optimally. Such a task is represented by an //objective function//which returns ​number that indicates how well a task is performed. In SCONE, objective functions are defined as a weighted combination of so-called //​Measures//;​ examples of which are walking speed, energy expenditure,​ center-of-mass position, etc. +
-</​panel>​ +
- +
-<panel title="​Optimizer">​ +
-The goal of an Optimizer is to find the parameters for which an objective function is minimized (or maximized, depending ​on the type of objective). SCONE uses a //shooting-based// approach to optimization,​ in which a simulation is performed multiple times, and parameters are adjusted according the result of the objective function +
-</panel> +
-</​accordion>​ +
- +
-====== Who is SCONE for? ====== +
-SCONE is designed for biomechanics,​ robotics and neuromechanics researchers or enthusiasts who wish to use predictive simulations in their research. For more information,​ see the [[faq]].+