ARI'S SAFETY PATENTS CAN SAVE LIVES
ARIhas developed, and is expanding, a strong IP position in autonomous highway safety. Currently, we have published applications and 7 issued Patents - with numerous applications before the USPTO for life-saving harm avoidance strategies, including novel V2X, AI, and 5G/6G algorithms.
ISSUED UNITED STATES PATENTS and PUBLISHED APPLICATIONS (click number to see the full document or click here to download the zip file of Patents & Publications as of 8/01/22):
V2X and Vehicle Localization by Local Map Exchange in 5G/6G - May 19, 2022
Vehicles in traffic can avoid accidents by coordinated avoidance, but this requires that they know which vehicle has which wireless address. In this patent, each vehicle prepares a small map fragment by measuring angles between each pair of the surrounding cars. They exchange the geometrical data, and then one of them "merges" the fragments to create a full traffic map.
Result: EACH VEHICLE KNOWS WHERE EACH OTHER VEHICLE IS POSITIONED.
Vehicle Connectivity, V2X Communication, and 5G/6G Sidelink Messaging - May 12, 2022
Normally, there is no way for a vehicle to know the wireless address of another vehicle in traffic. In this patent, each vehicle displays a "connectivity matrix", which is a sign or label that indicates the wireless address in a computer-readable pattern. After reading the code, each vehicle can communicate with the others to avoid collisions and smooth the flow of traffic.
Result: VEHICLES COOPERATE FOR TRAFFIC MANAGEMENT AND COLLISION AVOIDANCE.
V2X Messaging in 5G/6G with Simultaneous GPS Reception - May 12, 2022
Vehicles in motion have difficulty obtaining sufficient resolution from GPS, due to numerous motion-based distortions. In this patent, superior resolution can be obtained, and many sources of error can be canceled, by arranging for the vehicles to all acquire the same satellite signals at the same time. They then compare signals to get a differential position of each vehicle.
Result: HIGH-PRECISION LOCALIZATION OF EACH VEHICLE.
Autonomous Vehicle Localization System - October 27, 2020
In this patent, as soon as vehicles come into range with each other, they emit pulses that encode their wireless addresses. Other vehicles can receive the data optically. After exchanging addresses, they can then communicate directly with each other.
Result: COORDINATION FOR SAFETY AND TRAFFIC FLOW.
Autonomous Vehicle Localization System - October 27, 2020
To coordinate their actions, vehicles in traffic need to localize each other. In this patent, each vehicle emits a pulse of light and simultaneously transmits a wireless message, thereby identifying themselves to other vehicles. They can then communicate, especially in an imminent collision.
Result: COOPERATIVE COLLISION AVOIDANCE.
Rapid Wireless Communication for Vehicle Collision Mitigation - July 14, 2020
In an emergency, the computer on a vehicle may not be able to find the best avoidance maneuver in time to avoid the collision. This patent shows how vehicles can get help from a land-based supercomputer, which explores millions of avoidance scenarios (including pre-tested ones), selects the most effective, and transmits it back to the vehicle in milliseconds.
Result: VEHICLE HAS THE BENEFIT OF BEST KNOWN SOLUTION.
Systems and Methods for Hazard Mitigation - December 17, 2019
This patent discloses a learning-based collision-mitigation system, adaptable to the driver's own preferences and driving style. In an emergency, the processor compares previously-successful avoidance strategies in the current traffic scenario, and then quickly puts the most beneficial strategy into action. The method is applicable to autonomous vehicles, semi-autonomous, and human-driven with an emergency-intervention capability.
Result: BETTER COLLISION AVOIDANCE THAN HUMANLY POSSIBLE.
Systems and Methods for Hazard Mitigation - August 28, 2018
Unlike prior-art systems, this patent discloses ways to actively manage an unavoidable collision. By following a pre-planned strategy, each vehicle can shed excess kinetic energy, carefully place the point of impact away from passengers, and then swerve if necessary to avoid a spin-out. The system also plans ahead to avoid post-collision hazards such as secondary collisions.
Result: EVERYONE WALKS AWAY.
Systems and Methods for Hazard Mitigation - February 20, 2018
Most collision-avoidance systems simply slam on the brakes and hope. That is seldom the best solution. In this patent, we disclose a method for avoiding a collision if avoidable, and minimizing the harm if unavoidable. Unlike prior art, the amount of expected harm is calculated according to objective criteria.
Result: SAVING LIVES ON THE HIGHWAY.
Systems and Methods for Hazard Mitigation - July 11, 2017
Each year 35,000 people die on US highways - needlessly. This patent discloses procedures and a formula for mitigating imminent collisions. Detect the imminent collision, determine if it is avoidable, and execute a calculated evasive action. If not avoidable, select a different action to minimize the harm. The processor explores a huge number of alternative solutions and selects the best one for the desired outcome in each case.
Result: FEWER COLLISIONS.
ARI Patents are focused on autonomous collision mitigation. This entails systems and methods for determining whether an imminent collision is avoidable, planning a defensive strategy based on the vehicle's maximum braking, steering, and acceleration capabilities, and then carrying out that strategy. The mission is to avoid a collision if it is avoidable, manage the collision to minimize harm if it is unavoidable, and have the instant ability to know the difference.
Featuring a constantly updated kinetic model of the surrounding traffic, surprises by anything that happens are eliminated. Whenever a vehicle looks like a threat, ARI’s patented strategies extrapolate the model forward in time, exploring a wide range of actions to prevent the collision. Usually this is sufficient.
The ARI Advanced Collision Mitigation strategies combine lightning-fast collision avoidance with precise harm minimization - autonomously.
Unfortunately, a collision is sometimes unavoidable. In that case, the problem can be recognized early, and the overall harm can be evaluated – including major and minor injuries, and property damage likely from the collision if nothing is done. Then, in milliseconds with ultra high-speed processors, the likely effect of different courses of action can be tested, such as braking, accelerating, or steering at critical moments, and then the best strategy to minimize harm is then selected. The system then takes over and precisely guides the vehicle according to the least-harm trajectory. It continues to control the action all the way through the collision, millisecond by millisecond, doing everything possible to keep people safe.
A key feature of ARI Patents is a coherent post-collision strategy. The processors automatically develop a plan for avoiding secondary threats even before the collision occurs. During and after the collision, the sensors continuously keep a lookout for approaching traffic. The system drives the vehicle to the side of the road quickly and safely, then stops, at which point control can be turned back over to the driver. It even sends a help-request message with location and injury details.
Because the collision has been properly managed, it is far more likely that everyone walks away. The cars may be banged up, but serious injuries were minimized. Implementation of ARI strategies can save lives one more time.
Following is a brief list of some of the ARI patented strategies featuring:
- A system comprising processors configured to determine, from sensor data, one or more of a position, a velocity, and/or an acceleration of a subject vehicle; determine whether a collision between the subject vehicle and the second vehicle is imminent; calculate one or more sequences to avoid the imminent collision or to minimize harm of the collision, wherein each sequence comprises accelerations or decelerations or steering action; response to a determination that the collision is avoidable, select a sequence to avoid the collision; responsive to a determination that the collision is unavoidable, select a sequence to minimize the harm of the collision; and then implement the selected sequence by sending control signals to means for accelerating, decelerating, and steering the subject vehicle.
- A processor configured to determine, from sensor data, the position, velocity, and acceleration of a second vehicle, then determine, from the sensor data, whether a collision is imminent. The processor then determines whether a collision is avoidable by a particular sequence of accelerations, braking, or steering actions, and implements the particular sequence, when the collision is avoidable.
- If the collision is unavoidable, the processor calculates the harm associated with the collision, using a formula to quantify different types of harm. The processor then instantly selects a sequence of actions that minimizes the expected harm of the collision, and then implements that sequence.
- The system includes internal sensors that measure a position, a velocity, an acceleration, a deceleration, a steering status, or a steering action, of the subject vehicle, and external sensors that measure an image, a position, a velocity, an acceleration, or a deceleration of a second vehicle.
- A collision warning device comprising an acoustical signal generator, a light flasher, or a haptic vibrator, is activated when a collision is calculated to be imminent. The collision warning device renders, when a collision is imminent, information about a direction from which a second vehicle is approaching a subject vehicle. The collision warning device includes a voice-like speech generator configured to render the direction from which a second vehicle is approaching a subject vehicle.
- An adjustment device configured to modify a processor operation based on an input by a user, wherein the adjustment device may be set to a particular setting, such that intervention is withheld, after a collision is calculated to be imminent, for a user-selected time period, and then is implemented if the collision remains imminent.
- The intervention system includes user-selectable intervention threshold, wherein the system is configured to calculate a degree of hazard, and to implement a strategy if the degree of hazard exceeds the intervention threshold.
- A data storage module coupled to a processor and configured to store and protect critical data, comprising data related to traffic in a time period prior to a collision, data related to a collision, data related to the subject vehicle in a time period prior to a collision, and data related to any sequence of actions implemented prior to a collision.
- The system further including a data storage module which is hardened against damage caused by a collision, and against overwriting.
- Determining: if the subject vehicle and the second vehicle will pass within a predetermined radius of each other in the absence of alterations in the direction or velocity of the subject vehicle.
- Calculating: from the position and velocity and acceleration of the second vehicle, and from the position and velocity and acceleration of the subject vehicle, future values of a separation distance between the subject vehicle and the second vehicle; calculating from the future values a collision time at which the separation distance is less than a predetermined separation distance; and determining, if the collision time is less than a predetermined time limit, that the collision is imminent.
- While the particular sequence is being implemented, continuing to analyze further sensor data, thereby determining if the collision remains avoidable or unavoidable; if the continuing analysis indicates that an avoidable collision has become unavoidable, responsively implementing the particular sequence associated with the least harm; and if the continuing analysis indicates that an unavoidable collision has become avoidable, responsively implementing the particular sequence that avoids the collision.
- Receiving: capability-data comprising the maximum acceleration or deceleration or steering that the subject vehicle is capable of; and analyzing, with the capability data, whether the imminent collision can be avoided by applying the maximum acceleration or deceleration or steering to the subject vehicle.
- While the particular set of sequential actions are being implemented, preparing a feedback signal by comparing the measured position or velocity or acceleration of the subject vehicle to the particular set of sequential actions, and controlling the accelerator or brakes or steering of the subject vehicle according to the feedback signal in real time.
- Calculating harm: comprising calculating how many fatalities would result from a collision; calculating how many injuries would result from the collision; calculating how much property damage would result from the collision; adding, for each of the analyzed collisions, the calculated number of fatalities times a predetermined fatality coefficient, plus the calculated number of injuries times a predetermined injury coefficient, plus the calculated property damage times a predetermined property damage coefficient, wherein a sum of the adding indicates how much harm would be caused by a collision according to each of the sequences.
- Calculating harm including: predicting vehicle distortions that would occur during a possible collision; predicting peak accelerations that would occur during the possible collision; estimating, from the predicted vehicle distortions and peak accelerations, a number of fatalities, a number of injuries, and an amount of property damage that would result from the possible collision; and combining, according to a formula, the estimated number of fatalities, and the estimated number of injuries, and the estimated amount of property damage, thereby calculating the expected harm of the possible collision.
- Preparing: before a collision occurs, a post-collision strategy to minimize post-collision harm; acquiring, during or after the collision, further sensor data; updating, according to the further sensor data, the post-collision strategy; and then implementing the updated post-collision strategy.
- Implementing: a post-collision strategy comprising at least one of: turning off a fuel pump; unlocking doors; rolling down windows; driving to a side of a road; and transmitting a help-request message.
- Determining: after a collision occurs, whether a driver of the subject vehicle is responsive or nonresponsive; while the driver is nonresponsive, implementing the post-collision strategy; and while the driver is responsive, halting the post-collision strategy.
- During the implementation of an avoidance or harm-minimization strategy, redetermining whether the collision has changed from avoidable to unavoidable, or from unavoidable to avoidable.
ARI's Patents combine COLLISION AVOIDANCE and HARM MINIMIZATION.