Connected VehicleExperts Say!

Smart trails and connected cars

In his famous book, “Surely You’re Joking, Mr. Feynman”, Richard Feynman has documented experiments and outcomes with ants and their trails. It is perhaps counter-intuitive to analyze these outcomes in a journal for Connected Cars. However it may excite the reader if certain parallels can be drawn between the navigation systems of modern cars compared to navigation of ants in search for food.


Ant trails are characterized by the following key tenets

  • Ants mark their path with trails on a common ground
  •  Trails get reinforced when ants successively use the same path
  •  On successful return (characterized by success in food search) by a different path, ants leave
    different and stronger trails to differentiate and indicate success to other ants starting to look for
  • Initial trails get superseded by successful return trails (when these are different)
  • Trails are chemical-based and generally evaporate in around 30 mins
  • Navigation systems of connected cars have the following characteristics
  • Cars “mark” their routes with invisible trails on a common map
  • Trails get “reinforced” when multiple cars use the same route
  •  A “successful sojourn” (characterized by less time, distance, fatigue etc) on a different route is
    more recent data and indicates to other cars of an alternate viable route at the start of a journey
  •  Initial sojourns get superseded by successful sojourns (when these are different)
  •  Trails can be stored on cloud and can be analyzed over several months to fine-tune

Connectivity enables smartness in systems. For example, an ant considers a simple but smart logic to choose a path that maximizes its chance of successfully finding food (refer Figure 1). If its predecessor 1 went on this path, it will choose the same. However, if its predecessor(s) returned on a different path successfully with food, it will prioritize this path over the former and which will then get reinforced with newer trails. Similarly for a “smart, connected” car in Figure 2, the mantra is that it will follow its predecessor 1 who went on a given route, unless a predecessor 2 has been able to use a better alternate route (that is longer but with less congestion resulting in overall saving in time) in which case the car will prioritize the latter route over the former and reinforce this route by choosing it. Since the
route selection criteria between ants and cars are different, the preferred route choice in itself could be completely opposite (as shown by thicker trails in the Figures 1 and 2).
harman-article-T'wireModern city traffic administrators can utilize the smart trails for better traffic management. Let us consider that a city traffic administration gets a live map of its routes characterized by experiences of the Connected Cars. For example, routes have pre-fixed capacities but different vehicle densities at different times of the day. A live map updated minute-by-minute helps the traffic administrators to direct the traffic from a central monitoring and command center. Armed with the live maps, government agencies such as traffic coordinators can work on the traffic data and recommend route preferences to individual drivers. How will this work in a day to day scenario? Let us analyze the city traffic (shown in Figure 4) between points A and B via 2 different routes at different times of the day. For simplicity each of the blocks are considered as 2 by 2 kilometers. Distance via routes 1 and 2 both are 10 kilometers. However, traffic patterns vary during different times of the day. A part of the city is depicted below.

At different times of the day, observed times via Route 1 are:
6:00 am – Time (AB) = 4 + 4 + 4 + 4 + 4 = 20 mins
7:00 am – Time (AB) = 4 + 4 + 6 + 4 + 4 = 22 mins
8:00 am – Time (AB) = 4 + 6 + 6 + 4 + 4 = 24 mins
9:00 am – Time (AB) = 4 + 6 + 8 + 6 + 4 = 28 mins

Route 1 being a preferred route starts getting a higher vehicle density during the peak hours. Route 2 too
gets a higher vehicle density by around 9 am. Journey duration without the involvement of the traffic
coordinators are shown below.

Start Time Route 1 Route 2
6 am 20 20
7 am 22 20
8 am 24 21
9 am 28 22

Table 1: Typical Journey Duration in the Morning
Traffic coordinators can route vehicles via Route 2 to reduce the vehicle density on Route 1. With optimal
traffic flow, traffic coordinators can improve travel times on Route 1.

Start Time Route 1 Route 2
6 am 20 20
7 am 21 21
8 am 22 22
9 am 24 23

Table 2: Improved Journey Duration on Route 1 with Traffic Coordination

As depicted in the above chart, traffic coordinators are able to improve the travel times on Route 1 without significantly compromising the travel times on Route 2. There is an urgent need to research traffic flow and travel times by each of the city traffic administrators as every city may have different needs and peculiarities. As an example, a small industrial town may have peak hours only during a 30-60 minutes window twice a day, whereas in a metro there may be several peak hours during the day. Further, in an industrial town the traffic hotspot may be confined to a few pockets in the city – typically the entry and exit points into the factory. However, in a metro the traffic hotspots are spread out in several parts of the city. Moreover, travel time on a route may change over time in a day, week, month and year. Cyclical trends can emerge in daily, weekly, monthly and annual frequencies that
needs to be understood over a significant period of time and large volumes of data. For example weekend traffic patterns would be quite different than weekday traffic and summer traffic pattern could be quite different from the winter traffic pattern. Further, traffic data must be viewed with a “window”. More recent travel data (for e.g. time taken by a car) could have higher weightage compared to other travel data (for e.g. average times in the previous day or over last week). Besides, traffic data may need to be modulated for sudden events with no previous inkling for example a street protest.


In Emergency
We now consider other benefits of the Smart Trails in developing nations that often do not have sufficient width of roads to accommodate emergency vehicles on a dedicated lane. With burgeoning traffic in these countries, it is close to impossible to provide any prioritization to emergency vehicles. This often causes delays – sometimes fatal – in making an emergency services available. In certain other cases, entire stretches of roads are required to be closed to common traffic to enable emergency services such as for example organ transport. With connected cars, traffic coordinators can easily manipulate traffic on Route 1 in a way such that the journey duration is around 20 minutes and allows an emergency vehicle such as an ambulance to use Route 1 above even during peak hours and without delaying daily commuters on the route.

In other emergency situations when a “connected” car is directly involved, it would fare better compared to another vehicle as it can automatically “summon” help from Roadside Assistance for vehicle breakdown and Police or Ambulance for more severe situations e.g. an accident. In emergency situations on the road for e.g. a collision and pile-up, connected cars can (besides rerouting without dropping a sweat) provide vital information to traffic monitors even before people actually call for help. For example vehicles standing still in locations where they are unlikely to be standstill can surely trigger an action on traffic monitors to send in a patrol to evaluate the situation from up close.

City Parking
City parking is another concept that can be significantly improved by using Connected Car technologies. Many of the city roads allow parking on the weekends. For e.g. a street with multiple offices will most likely wear a deserted look on the weekends and can allow parking of hundreds of vehicles from the neighboring downtown shopping district. A connected car will have an advantage over traditional cars since it will possibly get notified for additional parking space available on the office street adjacent to the shopping district.
Further, city traffic administrators can find out the parking needs of a particular area by analyzing the smart trails and can proactively direct traffic towards the new parking plaza. While a connected car with a suitable parking database service will find the new plaza quite easily it may not be intuitive for other vehicles to look for this parking plaza. This is precisely where the traffic administrators will have a significant role to direct traffic to the new parking plaza until it is common knowledge and the parking plaza achieves the desired capacity. Similarly, traffic administrators can have better predictability of vehicle movement at entrances and exits of large malls.

Stolen Vehicle Tracking and Driver Behavior
These same trails that are used for betterment of traffic experience in a city are equally handy in tracing stolen vehicles. Law enforcers can utilize “connected car” trails to track the movement of stolen vehicles and nab the culprits before they cause much damage. In fact a Connected Car would be very unlikely to be stolen in the first place as an intrusion will possibly trigger an alarm and may force the thieves to abandon the plan and run for cover. Even if they do manage to take off with one such vehicle, they would soon be on the radar of the police and find themselves behind bars in no time.
Analysis of the same trails can also throw light on driver behavior. Imagine handing over your connected car to your young relative – either your teenage child or a sibling for an overnight trip by road. You can sleep easy if you get periodic updates from the car directly that it is safely driving within the speed limit and confirms on arrival at the destination. Moreover, after the trip one can analyze any risky driving pattern and eliminate the same with conscious intent. Safe driving habits are ultimately what will make roads safe.

Fuel/Energy Saving and Smart Home Connectivity
Connected cars can also enable energy and fuel saving in the city. Smart street lighting must be setup based on the traffic patterns to save cost and energy. For example, during the normal evening hours there would likely be a continuous stream of vehicles on the Route 2. However, later during the night, many parts of the street lighting can be switched off. Instead, only portions of the street – where there is active traffic or pedestrian movement – can be kept in on state while switching off portions where there is no traffic activity. Significant saving in fuel is also an outcome of “Smart” cars as they will spend far less time and distance to explore better routes or to explore availability of parking.

Now, consider another situation. You have setup an alarm to wake up at 7 am to wake up and get ready for the daily the commute to office for a schedule meeting at 9 am. You have planned to hit the road at 8 am. At 7:45 am, you get a message that the traffic has peaked early today and as a result if you hit the road at 8am the commute time is 15 minutes higher than other days. Instead you get a recommendation to start around 8:30 when it is predicted based on previous traffic patterns that the commute time is likely to be within 1 minute of the average commute time. What would you do? You would more than likely postpone that meeting at 9 am to 9:30 am and then start for office at 8:30 am rather than at 8 am. A connected car can thus improve the driving experience even before you hit the road. Once you hit the road, and are almost halfway down to office, you remember the Air Conditioning at home has not been switched off! Again your connected car comes to your rescue. The connected car connects to the Smart Home hub over internet and seamlessly switches off the Air Conditioning. On your way back from office, you could switch on the Air Conditioner for optimal temperature when you reach home.

Self-Learning and Autonomous Cars
In several countries – especially developed ones – we see a trend of Autonomous Driving and Self-Learning Cars. Some of the Autonomous cars can navigate on their own on specific traffic situations for example bumper to bumper traffic. The self-learning car can predict the route its driver will want to take based on past driving history. Most of these cars will also have Software Update capability with improved algorithms and bug fixes for its navigation or learning modules. As highlighted earlier, significant research is required in each of the areas noted above before a well-rounded experience is achieved with a combination of a fleet of Smart Connected vehicles as well as Analytics.

As the infotainment system in a connected Car has progressively become the fifth screen after TV, Computer, Tablet and Mobile, it also encompasses a payment gateway and convenient shopping hub. On your way to office you could pre-order and pre-pay for your coffee and pick it up on the way. On a return trip from office, you could pre-order your vehicle part and book an appointment with your service dealer. On a weekend trip you could delight your spouse by booking tickets to the concert and by locating a quaint little restaurant snuggled by the roadside off the highway.

Connected cars can delight the consumers with their “smartness” and enabling previously unthinkable use cases in addition to providing unmatched convenience and safety for the car and driver. We now consider an architecture (in Figure 5) of a connected car ecosystem that would enable this kind of “smartness”. As indicated, connectivity is a key recipe of “smartness”. Connectivity could be either “built-in” or “brought-in” through a smartphone. Besides, analytics is also an equally important ingredient as is personalization. The entire ecosystem needs to be hosted on the cloud and must have software update capability – similar to what we today have for Smartphones. Besides, applications and content approved by vehicle manufacturers must be available in the vehicle without compromising on safety and security.

In conclusion, while ant trails have been successful in explaining the social communication traits among ants, Connected Cars are essentially a manifestation of mankind’s social traits. Several aspects of how Connected Cars have a positive influence on friends and family members, other vehicle users and city administrators in the social chain is explained with Navigation, Traffic, Parking and Emergency use cases in cities. Connected Cars technology will continue to further evolve with a plethora of convenience and safety features.

The ability of the ”Connected Car” technology to delight the consumers shall spur its growth and adoption in vehicles. Both auto manufacturers and consumers are keen to explore the benefits that come with the Connected Cars. Further, Connected Cars is becoming an important link in the smart life of end users who have access to a variety of smart devices from smartphones to smart home hubs and demand similar or higher “smartness” from their vehicles. Auto manufacturers are game!


Related Articles