We are exploring the contributions of next-generation user interfaces in traffic stops.
Traffic stops
Traffic stop operations represent one of the most common interactions between police officers and civilians. They are often a low-risk routine but have the potential of risk escalation due to operational uncertainties that may trigger situations that can compromise both officer and civilian safety. These uncertainties are caused by the time gap between gathering the data about the vehicle and its occupants and verifying the received information. The officer needs to approach the driver, ask for documents and run the data in various databases to fully identify the situation. In some cases, the officer can be dealing with a criminal and only realize it when it is too late. Moreover, the officer might lose track of the vehicle’s occupants when he needs to use the in-car computer to verify the information.
VR Simulation
We created a traffic stop scene, with three procedures in mind: inspecting and approach the vehicle, interviewing the driver and checking documents, and proceeding with warnings, tickets, or severe/escalating actions. The proposed simulated AR UI is intended to support these tasks. The simulation starts with the suspect’s vehicle pulled over and the officer approaching it from behind, a common procedure that exposes the officer to a high degree of risk.
- Vehicle approach: While approaching the vehicle, the officer performs a visual inspection looking for suspicious movements and illegal objects inside the vehicle.
- Driver interview: The officer interacts with a virtual avatar. The avatar can understand natural language and will answer to predetermined questions. For example, the driver will hand their driver’s license upon request. Alternatively, a system control UI enable the interaction if voice commands are not possible or desirable.
- Information verification: Once the officer is in possession of the driver’s documents, the information about the driver is searched.
Next-generation User Interface
We conceived a next-generation AR UI to assist police officers in a traffic stop operation. The design leveraged knowledge that we gathered in the requirement analysis, leading us to prototype a system with two high-level features: a situational awareness interface and an on-demand information display. Used together, these features facilitate quick and easy access to information while minimizing the risks involved. Our designs envision technology that performs real-time scanning and searching of vehicle’s plate, driver’s license, facial recognition, and object detection, as it is reasonable to assume that such these are features will be supported by future technology.
Situational Awareness Interface
The situational awareness interface is designed to help the officer perceive key elements in the environment and to assess potential risk escalation. Simulated pattern recognition tags key elements, such as plates, driver’s license, and suspicious objects in the scene. An alert system provides a status update when the information searched in the background is ready. The alert system works at three distinct levels: low priority, medium priority, and high priority. Low priority alerts are informational. They notify that the data about a query was fetched and no irregularities were found. Medium priority alerts may require attention but are not time sensitive. For instance, this kind of alert may indicate an expired vehicle registration. With high priority alerts, however, immediate action must be taken as the situation may put the officer in danger. Normally, after a high priority alert, no further information is needed. Sufficient evidence is likely available to support a warrant for arrest (or to execute an existing one). The alert’s intensities have distinct encoding so it is easy to understand the severity of the alert. We support the delivery of alerts through the visual and haptic channels. Visual alerts are delivered as color-coded signals in the officer’s peripheral view, and vibrations are encoded by frequency and intensity to express each alert priority level.
On-demand Information Display
Using feedback gathered from law enforcement officers during the requirement analysis phase, we designed an on-demand information display that follows a physical armband mounted on the non-dominant forearm and is activated on demand through glance. This arm-mounted display interface is always within reach, yet readily removed from view without needing to divert attention from other high priority concerns. Both hands, therefore, remain free to perform other crucial tasks as new information is delivered and the traffic stop runs its process. This way, the officer can focus on the traffic stop procedures without being overwhelmed with information. The information display interfaces with the car’s computer system and provides a summary of relevant data about individuals and vehicles encountered during the traffic stop. Information is grouped into two categories. One for the vehicle, and another for the driver. It is possible to navigate through the categories by pressing virtual buttons on the display’s graphical user interface (GUI). The situational awareness interface alerts the officer when new data is available so that they can choose to verify the info through the arm-mounted display.
We designed and prototyped a 3D-printed tracked armband to properly track the forearm in the virtual space. The device is worn near the wrist, much like a wristwatch, by the hand where a phone or radio might also be held. As such, this is a natural place to look for information. Wrists are highly sensitive to vibrations, followed by the arms. Bony areas are best suited for detecting vibrotactile stimuli [13]. The armband’s haptic system uses two linear resonant actuators(LRA) vibration motors for alerts and notifications. Haptic alerts are adjustable for amplitude, frequency, and duration. With a low profile design and adjustable strap, rotational forces are effectively distributed while remaining in close contact with the wrist.
VR Traffic Stop simulation with Next generation UIs
The video shows the traffic stop procedure assisted by next generation UIs.
We designed the AR interface to assist in a specific stage of a traffic stop, where the officer already pulled over the suspect vehicle and is starting the approaching and interrogation of its occupants. The scenario also assumes that the officer has no prior knowledge about the situation. In circumstances, we can demonstrate all the benefits of the proposed interface. While pulling over a car and interrogation can happen without a prior check on the vehicle’s information, commonly the officer would run the vehicle’s plate before living the patrol vehicle. During the discussion with law enforcement agents, we prototyped an in-car interface that could assist the officer while patrolling the streets. While we acknowledge that an in-car UI would complete the traffic stop scenario experience, based on the discussions, we decided to focus on the second stage since it is where the agent is most vulnerable.