WheelCon is a novel, free and open-source platform to design video games that noninvasively simulates mountain biking down a steep, twisting, bumpy trail. It contains components presenting in human sensorimotor control (delay, quantization, noise, disturbance, and multiple feedback loops) and allows researchers to study the layered architecture in sensorimotor control.
Feedback control theory has been extensively implemented to theoretically model human sensorimotor control. However, experimental platforms capable of manipulating important components of multiple feedback loops lack development. This paper describes WheelCon, an open-source platform aimed at resolving such insufficiencies. Using only a computer, a standard display, and inexpensive gaming steering wheel equipped with a force feedback motor, WheelCon safely simulates the canonical sensorimotor task of riding a mountain bike down a steep, twisting, bumpy trail. The platform provides flexibility, as will be demonstrated in the demos provided, so that researchers may manipulate the disturbances, delay, and quantization (data rate) in the layered feedback loops, including a high-level advanced plan layer and a low-level delayed reflex layer. In this paper, we illustrate WheelCon's graphical user interface (GUI), the input and output of existing demos, and how to design new games. In addition, we present the basic feedback model and the experimental results from the demo games, which align well with the model's prediction. The WheelCon platform can be downloaded at https://github.com/Doyle-Lab/WheelCon. In short, the platform is featured to be cheap, simple to use, and flexible to program for effective sensorimotor neuroscience research and control engineering education.
The human sensorimotor control system is extremely robust1, although the sensing is distributed, variable, sparse, quantized, noisy and delayed2,3,4; the computing in the central nervous system is slow5,6,7; and the muscle actuation fatigues and saturates8. Many computational theoretical models have been proposed to explain the complicated human sensorimotor control process4,9,10,11,12,13,14, which is a tradeoff process in human reach and response15,16. For example, feedback control theory predicts the optimal control policy12, Bayesian theory models sensorimotor learning17,18,19 and information theory sensorimotor foundation20,21. In contrast to the abundance of theoretical models, experimental platforms capable of manipulating important components of multiple feedback loops lack development. This is in part due to the fact that designing a platform to bridge and test these aspects of sensorimotor control requires a diverse range of expertise, extending from motor control theory, signal processing, and interaction, all the way to computer graphics and programming. Researchers often develop their own custom hardware/software systems to characterize human sensorimotor control performance, which can limit the ability to compare/contrast and integrate datasets across research groups. The development of an easy-to-use and validated system could broaden the quantitative characterization of sensorimotor control.
In this paper, we present the WheelCon platform, a novel, free and open-source platform to design video games for a virtual environment that noninvasively simulates a Fitts’ Law reaching game and a mountain bike task with downing a steep, twisting and bumpy trail. The Fitts’ law for reaching task quantifies the tradeoff between speed and accuracy in which the time required for reaching a target of width at distance scales as22,23. The 'mountain-bike task' is a combination of a pursuit and compensatory tracking task, which are two classic components of research on human sensorimotor performance, especially in terms of studying feedback loops.
WheelCon contains the highly demanded basic components presented in each theory: delay, quantization, noise, disturbance, and multiple feedback loops. It is a potential tool for studying the following diverse questions in human sensorimotor control:
• How the human sensorimotor system deals with the delay and quantization in neural signaling, which is fundamentally constrained by the limited resources (such as the space and metabolic costs) in the brain24,25;
• How neural correlation in the human cortex with sensorimotor control26;
• How humans handle the unpredictable, external disturbances in sensorimotor control27;
• How the hierarchical control loops layered and integrated within human sensorimotor system16,28,29;
• The consequence of the delay and quantization in human visual feedback30 and reflex feedback31 in sensorimotor control;
• The optimal policy and strategy for sensorimotor learning under delay and quantization16,17,24,29.
WheelCon integrates with a steering wheel and can simulate game conditions that manipulate the variables in these questions, such as signaling delay, quantization, noise, and disturbance, while recording the dynamic control policy and system errors. It also allows researchers to study the layered architecture in sensorimotor control. In the example of riding a mountain bike, two control layers are involved in this task: the high-layer plan and the low-layer reflex. For visible disturbances (i.e., the trail), we plan before the disturbance arrives. For disturbances unknown in advance (i.e., small bumps), the control relies on delayed reflexes. Feedback control theory proposes that effective layered architectures can integrate the higher layers' goals, plans, decisions with the lower layers' sensing, reflex, and action24. WheelCon provides experimental tools to induce distinctive disturbances in the plan and reflex layers separately for testing such a layered architecture (Figure 1).
We provide a cheap, easy to use and flexible to program platform, WheelCon that bridges the gap between theoretical and experimental studies on neuroscience. To be specific, it can be used for examining the effects of delay, quantization, disturbance, potentially speed-accuracy tradeoffs. The variables that can be manipulated in control loops are shown in Table 1. It can also be applied for studying decision making and multiplexing ability across different control layers in human sensorimotor control. Moreover, WheelCon is compatible with noninvasive neural recordings, such as electroencephalography (EEG), to measure the neural response during sensorimotor control32,33,34,35, and the non-invasive brain stimulation techniques, such as Transcranial Electrical Stimulation (tES) and Transcranial Magnetic Stimulation (TMS), to manipulate the neural activity36,37.
The development and application of the protocol were approved by the California Institute of Technology Institutional Review Board (IRB) and the Southern University of Science and Technology IRB. The subject provided informed consent prior to any procedures being performed.
1. System preparation and setup
2. Task implementation
3. Data output
4. Input file development
Modelling Feedback Control
We show a simplified feedback control model shown in Figure 1. The system dynamics is given by:
where x(t) is the error at time t, r(t) is the trail disturbance w(t), is the bump disturbance, and u(t) is the control action.
Modeling Action Delay in Trail Disturbance
When there is a delay T in action, and a trail disturbance r(t), we model the control action by
The game starts with zero initial condition: x(0) = 0. The controller κ generates the control command u(t) using the full information on the histories of state, disturbance, and control input. Here, the net delay T is composed of the internal delays in the human sensorimotor feedback and the delays externally added. The control command is executed with delay T ≥ 0. Sensorimotor control in the risk-aware setting motivates the use of L1 optimal control, and as such, the goal is to verify the following robust control problem
This problem admits a simple and intuitive solution. The optimal cost is given by
This optimal cost is achieved by the worst-case control policy u(t + T) = −r(t), which yields
Modeling Action Quantization in Trail Disturbance
When the data rate, R, in the control loop is limited, the control action is generated by the following feedback loop with communication constraints,
where is a controller, and
is a quantizer with data rate R ≥ 1, i.e. S is a finite set of cardinality 2R. The disturbance r(t) is infinity-norm bound and without loss of generality,
. The worst-case state deviation is lower-bounded by
and the minimum control effort is given by
Measures of Error
To quantify the performance, we measured the infinity norm error , mean absolute error (MAE) and root mean square error (RMSE). The infinity norm is defined as the maximum of the absolute errors, where
Mean absolute error is calculated as follows
Root mean squared error is calculated as follows
Game 1: Visual Advanced Warning or Delay
Game 1 evaluates how the length of the look-ahead window (advanced warning/delay) affects sensorimotor control performance without being exposed to additional disturbances.
Game 1 lasts for 360 seconds and consists of one continuous "Trail", which reduces the amount of look-ahead every 30 seconds. The game begins with 1 s of advanced warning, and then decreases to 0.75 s, and then to 0.5 s. From there, the game decreases the look ahead by 0.1 s until a minimum of -0.4 s is reached. Positive delay, or negative advanced warning, means only the trail behind the player is visible.
An evolution of error dynamics of the player as the game progresses with 1 s advanced warning and 0.4 s delay were depicted in Figure 3a-3b separately. Both the plot display only the middle 20 seconds of each of the 30-second intervals to neglect the effects of the player adjusting to the new look-ahead window. The progression of the error in the blocks looks stable in the 1 s advanced warning setting while in the 0.4s delay setting, the error flips upside down during the progress. To quantify that effect in more detail, we evaluate L1-/L2-/L∞– norm for the error dynamics for every 20 s group corresponding to a delay level. Summarizing these calculations in a plot gives Figure 3c, which demonstrates how the players' error-norm does not change until the advanced warning reaches 0.5 s and then increases in an approximately linear fashion.
Game 2: Delay in Action Output
Unlike Game 1's external visual delay, Game 2 adds specific internal delay to the action output; in other words, the current control policy u(t) works at u(t + Tact) where Tact is the external delay in action. Game 2 lasts for 180 s. Adjusting Tact every 30 s, Tact starts at 0 s, and increments by 0.1 s until it reaches 0.4 s.
The effects of delay in action are shown in Figure 4. Similar to the vision delay, the error increases linearly with the delay, which is well in line with the prediction from theory in Eq(3).
Game 3 and Game 4: Quantization in Vision input and Action Output
Game 3 and Game 4 study the effects of quantization in vision input and action output, respectively. Each game is 210 s long, and the quantization changes every 30 s, with the data rate increasing from 1 to 7 bits. For example, when the Rvis is 1 in Game 3, the desired position (gray line in the gaming GUI) is presented either in the center-left or the center-right of the screen. When Rvis = n, the desired position can be presented in 2n possible locations on the screen. For Game 4, when Ract = 1, the player is either going left or right with one speed. When Ract is n, the player can steer the wheel to go left or right with 2n-1 speeds.
The effects of quantization (limited data rate) in the vision and action are shown in Figure 5. Consistent with the theory's prediction in Eq(6), sensorimotor control performance improves with higher data rates and reaches the optimal control performance when R is around 5.
Game 5: Bump and trail disturbance
Game 5 is designed to test the effects of bump and trail disturbances on human sensorimotor control. Game 5 consists of three scenarios:
a) "Bumps", tracking a constant trail subject despite torque disturbances on the wheel that mimic hitting bumps when riding a mountain bike;
b) "Trail", tracking a moving trail with random turns but without bumps;
c) "Trail with Bumps", tracking a moving trail with random turns and bumps.
Each scenario lasts for 60 s in the order (Bumps, Trail, Trail with Bumps) with a 5 s rest preceding each scenario. Furthermore, the disturbances and the trail during the isolated phases are duplicated in the combined "Trails with Bumps" phase, so that a proper performance comparison can be drawn between the separate tasks and the one where the player must multiplex. During the entire game, there is 1 s of advanced warning in vision input, no delay in action output, and a 10-bit data rate for both vision and action.
As the disturbance, we use a random, binary signal, whose amplitude is the maximum possible torque the motor of the steering wheel can exert. In every 100 ms, the torque switches between max positive and negative (100 or -100 for the wheel). A similar random binary switching controls the trail derivative. More specifically, the trail travels at a constant speed but randomly switches its direction such that it always stays in the screen range comfortably visible to the player. We adjusted the velocity of the trail on the screen such that the required steering wheel turning rate is approximately 75°/s. Figure 6 illustrates the 5 s snapshots of the error dynamics for each scenario during the game.
NOTE: Since this is used to study the sensorimotor control performance with limits of delay and data rate, we only analyzed the data after the subjects were trained for the task, and their performance became stable. The learning effects have been excluded from the data. Moreover, the feedback control model has not considered learning.
Figure 1: Basic block diagram for an experimental platform with subject and gaming wheel with a motor.
Each box is a component that communicates or computes and has potentially both delay and quantization, including within the game in G. The advance warning T is also implemented on a computer screen with vision. Please click here to view a larger version of this figure.
Figure 2: The user-graphic interface for WheelCon.
(a) the main menu; (b) the Fitt's Law Task menu; (c) the Mountain Bike Task menu; (d) the video game interface for Fitt's Law Task; (e) the video game interface for Mountain Bike Task. Please click here to view a larger version of this figure.
Figure 3: Adding delay in vision input during the mountain-bike task.
(a-b) The system dynamics with time for the session with 1 s advanced warning (a) and with 0.4 s delay (b). The black line and blue line are the trail position and the player position, respectively. The red line is the error dynamics. (c) Error increases with the increasing delay. The negative delay means advanced warning. Please click here to view a larger version of this figure.
Figure 4: Effects of external delay in action on performance.
The L∞ norm of error, MAE and RMSE increases with the increasing delay. Please click here to view a larger version of this figure.
Figure 5: Quantization in vision input (a) and action output (b).
The L∞ error, MAE and RMSE are shown in the blue, black and red line, respectively. Please click here to view a larger version of this figure.
Figure 6: Effects of bump and trail disturbance on human sensorimotor control.
(a) the error dynamics induced by bump disturbance; (b) the error dynamics induced by trail disturbance; (c) the error dynamics induced by bump and trail disturbance; (d) the overlayed error in bumps (blue), trails (red), Trails with bumps (green). The purple-empty and orange-filled stem plots indicate the timing and direction of the bump disturbances and trail disturbance, respectively. Note that both the wheel forces and the trail rates are square waves, and the stems indicate where these square waves switch (i.e., derivatives of the forces and rates). Please click here to view a larger version of this figure.
Notation | Variable | Unit | Constrains |
w(t) | Bump disturbance | n Newton | 0 ≤ w(t) ≤ 100 |
r(t) | Trail disturbance | 100 pixels | 0 ≤ r(t) ≤ 100; |
r(t) | w(t) | |||
Tvis | Vision advance warning/delay | second | -1 ≤ Tvis < 1 |
Tact | Action delay | second | 0 ≤ Tvis < ∞ |
Rvis | External data rate in vision input | bit | 1 ≤ Rvis ≤ 10 |
Ract | External data rate in the wheel output | bit | 1 ≤ Ract ≤ 10 |
Qvis | External quantizer on the vision input | 1 | Qvis = 2Rvis |
Qact | External quantizer on the wheel output | 1 | Qact = 2Rvis |
Table 1: The variables in control loops which can be manipulated.
Notation | Variable | Unit | |
t | Time | second | |
x(t) | Error dynamics, measured by the distance between the player’s position and the desired position | 100 pixels | |
u(t) | Control policy, measured by the deviated angle of the steering wheel | degree |
Table 2: The state variable and input signal in the dynamic system.
Platform | Open Source ? | Download Link | Function |
2D Virtual Environment (2DVE) | No | 1.Reach-to-grasp movements for arm rehabilitation after stroke; | |
ETH MIKE | No | 1.For the assessment of proprioceptive, motor and sensorimotor hand impairments | |
OpenSim | Yes | https://opensim.stanford.edu/ | 1. To study neuromuscular coordination; 2. Let users develop models of musculoskeletal structures and create dynamic simulations of movements. |
PneuGlove | No | 1.To train and evaluate finger individuation in children with hemiplegic cerebral palsy | |
VirtualEnaction | Yes | http://virtualenaction.gforge.inria.fr/ | 1. For systemic neuroscience simulation; 2. To verify the functional models of the brain; 3.To experiment complex survival behaviors |
VR-SPIRIT | No | 1.To improve predictive abilities in social scenarios. 2. Rehabilitation intensive training for paediatric patients |
|
WheelCon | Yes | https://github.com/Doyle-Lab/WheelCon | 1. To verify the Fitts' Law in human sensorimotor control; 2.To study the layered architecture in Sensorimotor control. |
Table 3: The list of some existing sensorimotor platforms.
In this paper, we have presented a free, open-source gaming platform, WheelCon, for studying the effects of delay, quantization, disturbance, and layered feedback loops in human sensorimotor control. We have shown the hardware, the software, and the GUI. The settings of a single sensorimotor control loop with delay and quantization have been implemented, which allows us to measure the effects of delay, quantization, and disturbance in sensorimotor control. The experimental results are well in line with the prediction from the feedback control theory.
The protocols provide a way to noninvasively manipulate external delays and limit the data rate in both vision inputs and action output, and to analyze the sensorimotor control performance. In the protocol, we ask the participants to play the game under several task scenarios that have been pre-defined in the platform. With these tasks, we verified the linear effect of delay (Figure 3 and Figure 4) and the nonlinear effect of quantization (Figure 5). These effects imply the optimization given the speed-accuracy tradeoff in human sensorimotor control. The protocol also allows us to study the layered feedback loops in the human sensorimotor system with a high-level advanced plan layer and a low-level delayed reflex layer (Figure 6).
In this protocol, it is crucial to train, but not over train, the participant in advance; otherwise, the learning effects or the tiredness of participants will impact the predictions of the model. The variability of the sensorimotor ability across participants is inevitable, and therefore some parameters in the protocol (i.e., the wheel sensitivity and the torque of bump) need to be tuned based on participants' age, strength, and motor skills. Matching these parameters across groups is necessary. Here, we suggest that the user chooses an appropriate sensitivity for both the old and young group in order to make a comparison.
One limitation of the method is that the model presented here did not consider the learning process. It is important to note that we only analyzed the data after the subjects were well trained, and their performance became stable to avoid the learning effects.
Nicolas Denoyelle et al. developed a platform (VirtualEnaction) for systemic neuroscience simulation38. VirtualEnaction can be used to validate functional models of the brain. Besides, some rehabilitation platform has been developed for performing sensorimotor tasks39,40,41,42,43. Scott L. Delp et al. developed an open-source software (OpenSim) to create and analyze dynamic simulations of Movement in rehabilitation science43. Marika Demers et al. proposed a 2D virtual environment for arm rehabilitation after stroke41. Zbytniewska Monika et al. design a robotic device for the assessment of hand sensorimotor impairments42. James V. McCall et al. proposed a platform for finger rehabilitation in children with hemiplegic cerebral palsy40. Niccolò Butti et al. develop a VR-based platform about social prediction improvement and rehabilitation intensive training for pediatric patients39. Table 3 listed the comparisons between these platforms with the platform presented here.
For the future research directions, the platform is compatible with noninvasive neural recordings technique (EEG) to measure the neural response during sensorimotor control. Investigating the mapping relationship between sensorimotor control and the EEG spatial-frequency signal, we might reveal the brain mechanism of sensorimotor control. It will be an important research question for understanding the human sensorimotor system. Moreover, most of the theory in this study is based on optimal control after learning. The stopping time for training is quite arbitrarily and empirically chosen in the study. It is therefore an important issue to evaluate whether participants hit a hard asymptote or a plateau44. Future studies could investigate the sensorimotor learning study using WheelCon to further test the asymptote/plateau and their possible explanations with sensorimotor learning theory.
The authors have nothing to disclose.
We thank Mr. Zhengyang Wang for reshaping the scripts, shooting and editing the video, and Mr. Ziyuan Ye for editing the video. This study got support from CIT Endowment & National Science Foundation (to JCD), Boswell fellowship (to QL) and High-level University Fund (No. G02386301, G02386401), Guangdong Natural Science Foundation Joint Fund (No. 2019A1515111038).