Tension sensing and cooperative control method for dragging water supply hose by fire-fighting robot

Abstract

The invention relates to the technical field of automatic water supply control of fire-fighting robots, and discloses a tension sensing and cooperative control method for dragging a water supply hose by a fire-fighting robot, which comprises the following steps of: 1, initializing a system, and configuring basic parameters of the hose, the type of a robot platform, a safety threshold, an operation mode and environment prior information; and 2, multi-source state real-time collection is carried out, and hose and robot state data are obtained through a tension sensor, a displacement sensor, an attitude sensor, a torque sensor, motor current and a hydraulic monitoring unit. According to the tension sensing and cooperative control method for dragging the water supply hose by the fire-fighting robot, on the basis of multi-dimensional comprehensive sensing, through collection of total factors such as tension, resistance, bending, torsion, clamping stagnation, door clamping and a hydraulic state, the limitation that estimation only depends on the length traditionally is thoroughly eliminated, and an accurate and reliable real-time state basis is provided for water supply operation of the fire-fighting robot.

Description

Technical Field

[0001] This invention relates to the field of automatic water supply control technology for fire-fighting robots, specifically a tension sensing and collaborative control method for a fire-fighting robot towing a water supply hose. Background Technology

[0002] Firefighting robots, as core equipment replacing firefighters in performing reconnaissance, firefighting, and rescue operations at dangerous fire scenes involving high temperatures, dense smoke, toxic substances, and explosive materials, have become an important force in modern fire emergency rescue. During continuous firefighting operations, firefighting robots need to tow large-diameter, long-length water supply hoses to achieve long-distance water supply. The stability and controllability of the hose towing directly determine the robot's operational safety, water supply continuity, and firefighting efficiency.

[0003] Currently, the control of fire-fighting robot hose towing mostly adopts passive hose release and manual remote control. It relies solely on the hose release length to estimate the status, which has problems such as single perception dimension, poor control coordination, and insufficient autonomous handling capability. In complex operation scenarios such as doorways, corners, gaps, thresholds, and corridors, the hose is prone to tension overload, bending and breakage, jamming in door clamps, entanglement and loosening, resulting in water supply interruption and equipment damage. At the same time, the existing control schemes are not adapted to different robot platforms such as tracked, wheeled, and humanoid robots, resulting in poor versatility and difficulty in meeting the needs of intelligent, unmanned, and highly reliable combat operations, which seriously restricts the effectiveness of fire-fighting robot rescue.

[0004] Based on this, it is necessary to propose a tension sensing and collaborative control method for fire-fighting robots towing water supply hoses. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention provides a tension sensing and collaborative control method for a fire-fighting robot towing a water supply hose, thus solving the problems mentioned in the background technology.

[0006] This invention provides the following technical solution: a tension sensing and collaborative control method for a fire-fighting robot towing a water supply hose, comprising the following steps:

[0007] Step 1: System initialization, configuring basic parameters of the hose, robot platform type, safety threshold, operation mode and environmental prior information;

[0008] Step 2: Real-time multi-source status acquisition, using tension sensors, displacement sensors, attitude sensors, torque sensors, motor current, and hydraulic monitoring units to acquire status data of the hose and robot;

[0009] Step 3: Data preprocessing and state estimation. The collected data is filtered, fused, and calibrated to calculate the real-time tension, drag resistance, bending radius, torsion angle, hose length, and relative displacement of the hose.

[0010] Step 4: Multi-dimensional risk level determination. Based on the state estimation results, calculate and classify the risks of bending, jamming, door clamping, entanglement, overload, and interface loosening.

[0011] Step 5: Conventional collaborative control, under low-risk and medium-risk conditions, dynamically adjust the robot's speed, turning radius, walking posture, and obstacle-crossing strategy, and synchronously match the hose delivery / retrieval speed of the hose extension / retrieval mechanism;

[0012] Step Six: High Risk and Abnormal Handling. When a high risk or abnormal state is determined, perform actions such as deceleration, pausing, reversing, straightening, detouring, partial replanning, lifting the pipe, and yielding the pipe to actively resolve the blockage and abnormality.

[0013] Step 7: Fault reporting and mode switching. If autonomous handling fails, immediately lock the dangerous state, cut off the water supply or reduce the pressure, report the fault code and location information, and switch to manual remote control mode.

[0014] Step 8: Operation Resumption. Once the risk is eliminated or the anomaly is resolved, the system will automatically restart the operation process and restore normal towing water supply.

[0015] Preferably, the real-time acquisition of multi-source status in step two includes direct acquisition and indirect acquisition methods: direct acquisition is achieved through tension sensors, displacement sensors, attitude sensors, and hydraulic monitoring units; indirect acquisition is achieved through drive motor current, winch output torque, robot joint torque, and chassis load changes; the acquired data includes hose tension, robot motion parameters, hose geometric parameters, environmental contact status, and water supply pressure and flow rate.

[0016] Preferably, the equivalent tension in step three adopts a weighted fusion algorithm: The total resistance is modeled using a component-based composite model: .

[0017] Preferably, the risk level determination in step four adopts a multi-factor weighted assessment model: The risk level is assessed using four levels: low, medium, high, and emergency.

[0018] Preferably, in step five, the conventional collaborative control adopts an adaptive velocity law based on the risk level and platform type: .

[0019] Preferably, in step four, the anti-bending and anti-entanglement control involves real-time calculation of the bending radius. With bending factor ;when Or it may reverse the cumulative over-limit, triggering rollback, release compensation, and attitude correction.

[0020] Preferably, the amount of rollback in step six .

[0021] Preferably, the release and retraction mechanism in step five employs PID tension closed-loop control: .

[0022] Preferably, the cross-platform compatibility coefficient It is compatible with tracked, wheeled, and humanoid robots, and includes safety tension range, speed limit, turning radius, and obstacle crossing strategy.

[0023] Preferably, if the autonomous handling fails in step seven, the fault is locked, water supply is cut off, the issue is reported, and manual intervention is initiated; this satisfies... , , The job will automatically resume when needed.

[0024] Compared with the prior art, the present invention has the following beneficial effects:

[0025] 1. This tension sensing and collaborative control method for fire-fighting robots towing water supply hoses is based on multi-dimensional comprehensive sensing. By collecting all elements such as tension, resistance, bending, torsion, jamming, door clamping, and hydraulic status, it completely breaks away from the limitations of traditional methods that rely solely on length estimation. It provides accurate and reliable real-time status data for fire-fighting robots' water supply operations. At the same time, it constructs a complete intelligent control closed loop of perception, judgment, collaboration, handling, and recovery, realizing an upgrade from passive following to active prediction, hierarchical collaboration, and autonomous handling. Combined with four-level risk assessment and nine types of abnormal protection, it significantly improves the level of intelligent operation.

[0026] 2. The tension sensing and collaborative control method of this fire-fighting robot towing water supply hose has strong scene adaptability and platform versatility. It can cover complex indoor scenes such as doors, corners, gaps, thresholds, corridors, and oily floors, and supports continuous operation across rooms. It has a built-in parameter library for tracked, wheeled, and humanoid robots, which can be quickly ported and deployed without hardware modification and is compatible with different types of fire-fighting robot platforms. Attached Figure Description

[0027] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0028] Figure 1 This is a schematic diagram of the method flow of the present invention;

[0029] Figure 2 This is a schematic diagram of the tension sensing and resistance estimation of the present invention;

[0030] Figure 3 This is a schematic diagram of the closed-loop collaborative control of the present invention. Detailed Implementation

[0031] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0032] Please see Figure 1-3 A method for tension sensing and collaborative control of a fire-fighting robot towing a water supply hose includes the following steps:

[0033] Step 1: System initialization, configuring basic hose parameters, robot platform type, safety thresholds, operating mode, and prior environmental information.

[0034] The initialization parameters include the minimum allowable bending radius of the hose: Target tension: High and low tension thresholds: / Rated water supply pressure: Coefficient of friction: Mass of hose per unit length: Chief Steward: The weighting coefficients and the initial values ​​of the PID parameters;

[0035] Step Two: Real-time multi-source status acquisition. Status data of the hose and robot are obtained through tension sensors, displacement sensors, attitude sensors, torque sensors, motor current sensors, and a hydraulic monitoring unit.

[0036] The acquired data undergoes IP67-level sensor protection and hardware filtering, with a single frame acquisition cycle of [missing information]. =10-50ms, abnormal data is removed using the 3σ criterion to ensure the reliability of perception in harsh fire protection environments;

[0037] Step 3: Data preprocessing and state estimation. The collected data is filtered, fused, and calibrated to calculate the real-time tension, drag resistance, bending radius, torsion angle, hose length, and relative displacement of the hose.

[0038] Step 4: Multi-dimensional risk level determination. Based on the state estimation results, calculate and classify the risks of bending, jamming, door clamping, entanglement, overload, and interface loosening.

[0039] Step 5: Conventional collaborative control, under low-risk and medium-risk conditions, dynamically adjust the robot's speed, turning radius, walking posture, and obstacle-crossing strategy, and synchronously match the hose delivery / retrieval speed of the hose extension / retrieval mechanism;

[0040] Step Six: High Risk and Abnormal Handling. When a high risk or abnormal state is determined, perform actions such as deceleration, pausing, reversing, straightening, detouring, partial replanning, lifting the pipe, and yielding the pipe to actively resolve the blockage and abnormality.

[0041] Step 7: Fault Reporting and Mode Switching. If autonomous handling fails, immediately lock the dangerous status, cut off the water supply or reduce the pressure, report the fault code and location information, and switch to manual remote control mode.

[0042] The water outage employs a tiered pressure reduction and delay protection system, first reducing the pressure to 50%. Hold for 3 seconds before shutting off the water supply to prevent water hammer from impacting the pipeline;

[0043] Step 8: Operation Resumption. Once the risk is eliminated or the anomaly is resolved, the system will automatically restart the operation process and restore normal towing water supply.

[0044] Automatic recovery requires the following verification conditions: normal sensor, stable water supply pressure, no overturning or jamming / entanglement of the robot. Recovery is only possible if all conditions are met to prevent secondary failures.

[0045] As a preferred technical solution of the present invention, the multi-source real-time acquisition in step two includes direct acquisition and indirect acquisition methods: direct acquisition is achieved through tension sensors, displacement sensors, attitude sensors, and hydraulic monitoring units; indirect acquisition is achieved through drive motor current, winch output torque, robot joint torque, and chassis load changes; the acquired data includes hose tension, robot motion parameters, hose geometric parameters, environmental contact status, and water supply pressure and flow rate.

[0046] As a preferred embodiment of the present invention, the equivalent tension in step three adopts a weighted fusion algorithm: The total resistance is modeled using a component-based composite model: ,

[0047] In the formula: For equivalent comprehensive tension; For direct tension measurement by the sensor; Calculate tension based on winch torque; For observing the tension of motor current; For inverting the tension of robot joint torques; For the weighting coefficients, satisfying By unifying and fusing multi-source sensing information, robustness in harsh environments can be improved.

[0048] The rate of change of tension is: In the formula: The rate of change of tension; To control the cycle, it is used to identify transient risks such as sudden changes, door jamming, and stagnation. Typical threshold values ​​are specified. A value >50 N / s is considered a transient anomaly.

[0049] In the formula: Total drag resistance; Ground friction resistance; Adds resistance to turning; For bending deformation resistance; For concentrating resistance; Environmental interference resistance (door frame / door gap).

[0050] Among them, frictional resistance: In the formula: The coefficient of friction of the ground; Mass per unit length of hose; This represents the length of the pipe already laid. This refers to the total length of the hose; The acceleration due to gravity is taken as 9.8 m / s², which increases linearly with the length of the pipe, reflecting the drag loss on the ground.

[0051] The additional resistance during turning: In the formula: This refers to the steering angle; For robot speed; The turning radius; The lateral drag coefficient indicates that drag increases sharply during small-radius turns. When the resistance increases by ≥200%,

[0052] Among them, the resistance to bending deformation is: In the formula: is the bending stiffness coefficient of the hose; is the jamming resistance. Sewing with localized force abrupt changes; environmental interference resistance Observed from the contact force of the door gap / step.

[0053] As a preferred embodiment of the present invention, the risk level determination in step four adopts a multi-factor weighted evaluation model: The risk is assessed using four levels: low, medium, high, and emergency.

[0054] In the formula: Let the weights be the weights, and let the weights satisfy the following conditions: Typical values , , , , It can be dynamically adjusted according to the material of the hose; For pressure deviation; For the rated water supply pressure, output a continuous value of 0–1, mapping four levels of risk: low risk: 0–0.3, medium risk: 0.31–0.6, high risk: 0.61–0.85, and emergency risk: 0.86–1.0. The risk threshold can be dynamically and adaptively adjusted according to the ground material, hose diameter, and operating scenario.

[0055] As a preferred embodiment of the present invention, in step five, the conventional collaborative control adopts an adaptive velocity law based on the risk level and platform type: ,

[0056] In the formula: To set the speed; Output speed is determined by platform compatibility factors (different for tracked / wheeled / humanoid models). Set a lower limit Upper limit In addition, a first-order filter is added to limit the amplitude to avoid sudden stops and starts. The higher the risk, the lower the speed. In an emergency, the speed is set directly to 0.

[0057] As a preferred embodiment of the present invention, step four involves anti-bending and anti-entanglement control: real-time calculation of the bending radius. With bending factor ;when Or, the cumulative overshoot could trigger a rollback, pipe release compensation, or attitude correction.

[0058] In the formula: The radius of curvature; The arc length of the curved section; The change in bending angle is used to determine whether it is less than the minimum allowable radius of the hose. ,

[0059] In the formula: For bending risk factors; This indicates that the actual bending radius is smaller than the allowable radius, and the hose is at risk of bending damage.

[0060] Cumulative torsion angle: In the formula: The cumulative torsion angle is denoted as ; the single-step torsion increment is denoted as . Exceeding the threshold is considered a risk of entanglement, and the cumulative torsion angle is determined accordingly. The risk of entanglement was identified, triggering pipe release compensation and attitude correction.

[0061] As a preferred embodiment of the present invention, the amount of backtracking in step six ,

[0062] In the formula: This is the rollback ratio coefficient, typical value. Set an upper limit for rollback amount. rollback speed It also activates rear obstacle detection to avoid collisions. The priority of high-risk handling actions is: pause - back off - lift pipe - detour - replan. When multiple abnormalities are triggered, they are executed according to priority.

[0063] As a preferred embodiment of the present invention, the release and retraction mechanism in step five employs PID tension closed-loop control: ,

[0064] In the formula: This refers to the winch rotation speed; For target tension; Add a derivative term to the PID parameters. Improve dynamic response, enhance anti-integral saturation, output limiting and dead zone compensation, maintain constant tension, and avoid overstretching or relaxation.

[0065] As a preferred technical solution of the present invention, the cross-platform adaptation coefficient It is compatible with tracked, wheeled, and humanoid robots, and includes safety tension range, speed limits, turning radius, and obstacle crossing strategies.

[0066] tracked Wheel , humanoid Each corresponds to different obstacle-crossing capabilities and steering constraints.

[0067] As a preferred embodiment of the present invention, in step seven, if the autonomous handling fails, the fault is locked, the water supply is cut off, the issue is reported, and manual intervention is switched off; thus satisfying the requirements of the present invention. , , The job will automatically resume when needed.

[0068] Example 1: Tracked firefighting robot automatically supplies water to rooms and doors indoors.

[0069] Application scenario: In a residential building fire, the robot needs to enter the bedroom from the living room to extinguish the fire. It must pass through the bedroom door, where the floor is made of ordinary tiles, posing a risk of the robot getting caught in the door frame or gap.

[0070] Platform type: Tracked firefighting robot, weight 80kg, rated flow rate 30L / min, hose diameter 65mm, length 20m.

[0071] Release and retrieval mechanism: An electric winch release and retrieval mechanism is integrated at the tail of the robot.

[0072] Control method execution steps: Initialization: Configure the safe tension range as 200N–800N, minimum bending radius as 150mm, platform type as tracked, and door passage mode enabled; Status acquisition: Real-time acquisition of tension sensor data, winch motor current, robot speed, turning angle, attitude angle, and door frame approach signal; Status estimation: Fusion of tension sensor and motor current data to calculate equivalent tension; Calculation of ground friction resistance based on motion status; Risk assessment: As the robot approaches the door frame and the tension slowly increases, it is assessed as medium risk, and a door clamp warning is issued; Cooperative control: The robot… The robot's speed decreased from 0.8 m / s to 0.3 m / s, the winch started to deliver the hose, compensating for the length of the hose needed to pass through the door frame; Anomaly detection: the tension suddenly increased to 900 N the instant it passed through the door, indicating a high risk of being caught in the door; Autonomous handling: the robot immediately paused, retreated 150 mm, the winch continued to deliver the hose, and the robot's posture was slightly adjusted; Status recovery: the tension dropped to 500 N, the risk of being caught in the door was eliminated, the robot returned to normal speed, and continued to enter the bedroom to extinguish the fire; Operation maintenance: after entering the bedroom, the system maintained a low-risk state, provided a stable water supply, and continued firefighting operations.

[0073] Example 2: Automatic water supply operation for wheeled firefighting robot turning in corridors

[0074] Application scenario: In an office building fire, the robot operates at the bend of an L-shaped corridor, where flexible hoses are prone to lateral dragging, bending, and twisting.

[0075] Platform type: Four-wheel drive firefighting robot, with a small turning radius and flexible movement.

[0076] Release and retrieval mechanism: The base station side releases the pipe, and the robot's front end is equipped with a guide component.

[0077] Control method execution steps: Initialization: Set the safety tension to 300N–1000N, minimum bending radius to 200mm, and enable torsion monitoring. Status acquisition: Collect winch torque, robot position, turning angle, bending radius, torsion angle, flow rate, and pressure. Status estimation: Calculate additional turning resistance, lateral drag resistance, and changes in bending radius. Risk assessment: If the bending radius drops to 180mm during the turning process and the cumulative torsion angle exceeds the limit, the system is assessed as medium risk. Cooperative control: Reduce turning speed, increase turning radius, increase pipe laying speed at the base station, and release torsion stress. Status monitoring: Resistance decreases steadily, bending radius recovers to 250mm, and torsion angle decreases. Operation recovery: After the turning is completed, the system resumes normal operation and maintains stable water supply.

[0078] Implementation 3: Humanoid firefighting robot crosses threshold to lift and allow pipes to pass.

[0079] Application scenario: In old houses with high thresholds, flexible hoses are easily caught, jammed, or unable to be dragged.

[0080] Platform type: Bipedal humanoid firefighting robot, equipped with arm-lifting pipe lifting, posture adjustment, and balanced walking capabilities.

[0081] Deployment and retraction mechanism: Integrated deployment and retraction mechanism on the back, with arm-assisted guidance.

[0082] Control method execution steps: Initialization: Activate the hose lifting and yielding mode, set the threshold height to 100mm, and the minimum bending radius to 150mm; Status acquisition: Joint torque, posture sensor, visual threshold detection, tension sensor; Status estimation: Determine the contact state between the hose and the threshold, and calculate the jamming resistance; Risk assessment: Sudden increase in resistance, hose caught on the threshold, high risk is determined; Cooperative control: The robot stops moving forward, raises its arm to lift the hose over the threshold, the retraction mechanism micro-moves to deliver the hose, and adjusts its posture to pass smoothly; Abnormal resolution: The hose detaches from the threshold, the tension returns to normal, and the robot continues to move forward.

[0083] Example 4: High-risk emergency state fault protection

[0084] Application scenario: The hose is completely clamped by the door and cannot be released by its own retraction, resulting in a sudden drop in flow and an increase in pressure.

[0085] Control method execution steps: Risk assessment: When tension exceeds the maximum value, flow rate approaches zero, and pressure rises sharply, an emergency risk is assessed; Emergency handling: The robot immediately stops, the pipe-retracting mechanism is activated to slightly retract the pipe, and the water supply pump unit is linked to reduce pressure and stop; Fault reporting: Upload fault codes for door clamp jamming and flow interruption, and switch to manual remote control mode; Safety protection: Lock the mechanism to prevent accidental action and wait for manual handling.

[0086] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A method for tension sensing and collaborative control of a fire-fighting robot towing a water supply hose, characterized in that, Includes the following steps: Step 1: System initialization, configuring basic parameters of the hose, robot platform type, safety threshold, operation mode and environmental prior information; Step 2: Real-time multi-source status acquisition, using tension sensors, displacement sensors, attitude sensors, torque sensors, motor current, and hydraulic monitoring units to acquire status data of the hose and robot; Step 3: Data preprocessing and state estimation. The collected data is filtered, fused, and calibrated to calculate the real-time tension, drag resistance, bending radius, torsion angle, hose length, and relative displacement of the hose. Step 4: Multi-dimensional risk level determination. Based on the state estimation results, calculate and classify the risks of bending, jamming, door clamping, entanglement, overload, and interface loosening. Step 5: Conventional collaborative control, under low-risk and medium-risk conditions, dynamically adjust the robot's speed, turning radius, walking posture, and obstacle-crossing strategy, and synchronously match the hose delivery / retrieval speed of the hose extension / retrieval mechanism; Step Six: High Risk and Abnormal Handling. When a high risk or abnormal state is determined, perform actions such as deceleration, pausing, reversing, straightening, detouring, partial replanning, lifting the pipe, and yielding the pipe to actively resolve the blockage and abnormality. Step 7: Fault reporting and mode switching. If autonomous handling fails, immediately lock the dangerous state, cut off the water supply or reduce the pressure, report the fault code and location information, and switch to manual remote control mode. Step 8: Operation Resumption. Once the risk is eliminated or the anomaly is resolved, the system will automatically restart the operation process and restore normal towing water supply.

2. The tension sensing and collaborative control method for a fire-fighting robot towing a water supply hose according to claim 1, characterized in that: The real-time acquisition of multi-source status in step two includes direct acquisition and indirect acquisition methods: direct acquisition is achieved through tension sensors, displacement sensors, attitude sensors, and hydraulic monitoring units; indirect acquisition is achieved through drive motor current, winch output torque, robot joint torque, and chassis load changes; the acquired data includes hose tension, robot motion parameters, hose geometric parameters, environmental contact status, and water supply pressure and flow rate.

3. The tension sensing and collaborative control method for a fire-fighting robot towing a water supply hose according to claim 1, characterized in that: In step three, the equivalent tension is achieved using a weighted fusion algorithm: The total resistance is modeled using a component-based composite model: .

4. The tension sensing and collaborative control method for a fire-fighting robot towing a water supply hose according to claim 1, characterized in that: The risk level determination in step four adopts a multi-factor weighted assessment model: The risk level is assessed using four levels: low, medium, high, and emergency.

5. The tension sensing and collaborative control method for a fire-fighting robot towing a water supply hose according to claim 1, characterized in that: In step five, the conventional collaborative control employs an adaptive velocity law based on the risk level and platform type: .

6. The tension sensing and collaborative control method for a fire-fighting robot towing a water supply hose according to claim 1, characterized in that: Step four involves anti-bending and anti-entanglement control: real-time calculation of the bending radius. With bending factor ;when Or it may reverse the cumulative over-limit, triggering rollback, release compensation, and attitude correction.

7. The tension sensing and collaborative control method for a fire-fighting robot towing a water supply hose according to claim 1, characterized in that: The rollback amount in step six .

8. The tension sensing and collaborative control method for a fire-fighting robot towing a water supply hose according to claim 1, characterized in that: In step five, the release and retraction mechanism employs PID tension closed-loop control. .

9. The tension sensing and collaborative control method for a fire-fighting robot towing a water supply hose according to claim 1, characterized in that: The cross-platform compatibility coefficient It is compatible with tracked, wheeled, and humanoid robots, and includes safety tension range, speed limit, turning radius, and obstacle crossing strategy.

10. The tension sensing and collaborative control method for a fire-fighting robot towing a water supply hose according to claim 1, characterized in that: If the autonomous handling in step seven fails, the fault will be locked, water supply will be cut off, the incident will be reported, and manual intervention will be switched off; this satisfies the following requirements. , , The job will automatically resume when needed.

Images