Protecting our most precious asset … our children.
Introducing CAM! … the Crosswalk Assistant Machine (C.A.M.) of the future. CAM is a state-of-the-art social robotic crossing guard specially designed to assist elementary schools and pre-schools during school opening and closing times. CAM can skillfully manage any number of pedestrians at a crosswalk even in high traffic areas, making CAM an excellent resource for any urban school crosswalk.
CAM assists families and especially schoolchildren when crossing the street and helps them learn the safe way to do so. Stopping traffic from driving through the crosswalk is the main priority, therefore CAM uses multiple sensors that can detect cars within a 50 yard radius. When CAM detects pedestrians wishing to use the crosswalk, it will travel to the middle of the road and hold out a stop sign with flashing red lights and signal to the pedestrians that it is safe to cross the street once the traffic is safely stopped. When all have safely crossed the road CAM will then head back to the side of the road and wave traffic on.
CAM’s sensors actively monitor the environment for adverse weather conditions such as rain, snow, and ice. CAM’s algorithm logic will provide traffic with extra time to stop at the crosswalk in adverse conditions to improve safety of the pedestrians and other vehicles.
A continuous track system (like incorporated under a tank) is utilized so CAM can easily operate in adverse weather conditions and traverse curbs and other obstacles when entering and exiting the street. No crosswalk is left behind!
CAM uses state-of-the-art cameras and eye-safe laser range finding technology to detect and track cars, children, and the crosswalk for path planning. Add on camera features can also record traffic and the environment to supply potential traffic violations or dangerous activity to law enforcement.
Crossing guards have social personalities, therefore CAM is designed to be friendly and trusting, while also being authoritative to ensure traffic compliance. Humanoid features, such as large emotive eyes, are key. Average human height also ensures easy detection by cars. CAM implements machine learning techniques to “get to know” children and ask them about their day and school life. This social connection makes CAM a special friend for kids and families.
When there is no requirement for crosswalk services (overnight, school in session), then CAM “parks” itself in a predetermined location such as a nearby shack to charge it’s battery and await the next shift.
Background and Issues
Importance and Challenges
Streets near schools are often dangerous. Current safeguards to prevent schoolchildren from being hit by cars include lowering speed limits around schools and implementing crossing guards at peak hours. However, this puts the crossing guards at risk and assumes that they are alert at all times. With kids running around and cars coming from multiple directions, crossing guards might be subjected to a larger cognitive load than intended. This is coupled with environmental factors, such as the cold or rain, that might distract crossing guards and decrease the safety of the children crossing.
Directing traffic and predicting human driver behavior is challenging, and requires mutual understanding between the driver and the crossing guard. However, creating a safe environment around schools is important for parents, the children themselves, and for society as a whole.
To gain the requisite insight into the tasks that CAM would need to perform, our team interviewed crossing guards at local K-8 schools in suburban Boston and observed behaviors. We found crossing guards often have two shifts, five days a week from 7:45-8:30am and 2:30-3:15pm. They are challenged with tracking people crossing on all four cross walks when stationed at an intersection. They post at one corner, so people are encouraged to cross on the two crosswalks that the crossing guard has direct access to. However, children do not always comply with this convention and need extra attention. After noticing someone approach to cross, they look in all directions to confirm that traffic has stopped. They then shift their focus to the pedestrians.
Their typical interactions with children and parents include a general greeting (e.g. “Good morning, how are you?”) and an exchange of pleasantries with parents (e.g. “Thank you, have a good day”). They recognize some children specifically, who will sometimes initiate interaction themselves. Conversation is generally made while pedestrians are approaching the crosswalk and waiting and sometimes while crossing.
They had a generally positive reaction to the idea of a crossing guard robot and suggested adding a vest or stop sign to the design so its purpose will be recognizable.
Users range from children of all ages to adults and elderly. Young children will normally be escorted by their parents or guardians, however, some children might be crossing on their own or in a group. Children are less likely to monitor and understand their surroundings, and could miss warnings of unsafe situations. Children also have little to no understanding of traffic conventions and could misinterpret signals around them. Some users might have less mobility; disabled individuals and the elderly will require extra attention and time when crossing the street. Users might be moving with a variety of objects (strollers, bikes, etc.) and pets too. User’s might trip or drop items delaying the crossing time. All of these factors must be considered by CAM.
Parents desire the ability to ensure the safety of their children. They might be the most uncertain of this when their children are unsupervised, such as walking to and from school where they are more likely to encounter unsafe traffic, strangers, or bullies. They might also doubt the safety and efficiency of directing children in large groups when everyone gets out of school at the same time. In the end, parents aspire to teach their children proper safety precautions while avoiding causing a sense of paranoia and instilling a sense of confidence and independence in their children.
Kids desire independence as well and to prove they can be trusted to be safe and mature. However, they may be uncertain of busy, unfamiliar situations, such as traffic and the large groups surrounding their school at peak times. They might also doubt the authority of crossing guards or the idea that traffic may not stop for them, therefore rejecting the need for caution.
Select a user below to view their Persona!
Sensors, Processors & Actuators
CAM is loaded with high tech sensors, processors and actuators. Explore these features below!
Modes and Interactions
Select the Mode below to learn more
Select the Emotive Display below to learn more
Or scroll through the Task Analysis below!CAM-Task-Analysis
Real User Walkthrough
Frequently Asked Questions
* stop ahead
* a high volume of pedestrians using the crosswalk or populating the nearby area
* a stop that drivers often miss
* accidents or other hazards
Another feature that could be explored with future iterations of C.A.M is the implementation of two C.A.M robots in order to maximize the safety of those crossing the crosswalk. We envision having one C.A.M focus solely on incoming traffic while the other C.A.M is responsible for escorting the pedestrians to the other side of a crosswalk while being able to emotionally engage with the children/pedestrians.
A third feature that may be a bit more complicated to implement due to privacy considerations would be for C.A.M to monitor the adult walking with the child when they approach the crosswalk. C.A.M software would utilize machine learning facial recognition methods in order to recognize whether or not the person accompanying the child matches with people found in the database of “safe” or recognized adults. If C.A.M detects that this person is unknown, an SMS would immediately be sent to the parent or caregiver of the child in order to alert them. C.A.M would also have the ability to monitor dangerous or illegal scenarios in the environment. This could include speeding cars, drunk driving, as well as suspicious activity around schools. One of the limitations of monitoring or surveillance technology is public sentiment around losing privacy. This will make it more difficult for future iterations of C.A.M to include some of these features but it is very likely that society will grow accustomed (for greater or worse) to the increased use of surveillance and monitoring technology to ensure the safety of themselves and their loved ones.