Quantity turns into quality
Taking cues from nature for technical developments and solutions to problems is increasingly gaining ground in the concept of bionics – why reinvent a wheel that has been successfully spinning in nature for hundreds of thousands of years or even more? So it comes as no surprise that social phenomena in the animal kingdom are increasingly often the subject of study in terms of their emulation potential for technical processes as well.
When huge schools of herring glide through the water like one gigantic body that confuses attacking seals, or when bison gather around their young in a star-shaped formation – horns pointing outward – when threatened by a predator, we like to refer to this as swarm intelligence. This, in fact, is the same process behind such phenomena: Individuals, by behaving in a certain way, turn into a collective with characteristics that are beneficial to all but which cannot be produced by an individual. Is there anything to be learned from this?
Everyone acting in concert – does this already qualify as intelligent action?
Seti and Wikipedia are frequently cited as examples of swarm intelligence. In the case of Seti (Search for Extraterrestrial Intelligence) thousands of privately owned computers are connected in a joint search for indications of extraterrestrial intelligence in the signals that radio telescopes have recorded from the depths of the universe. And on Wikipedia, everyone can contribute articles to the world’s largest online encyclopedia or correct and/or expand existing ones.
It’s true that Wikipedia articles bear witness to the intelligence of their authors. Yet neither the swarm behavior of fish and bison, nor that of ET searchers or smart authors per se actually results in “intelligence” in the narrower sense. Therefore, synchronization behavior might be a more appropriate term. The animals instinctively react as a group and the collective participation in Wikipedia or Seti does not generate the type of quality that could exclusively be achieved this way but, first and foremost, serves to save time.
Seemingly meaningless data – important for the collective
The key to understanding genuine swarm intelligence is found in the question of whether or not coordinated behavior produces data which is useful to the collective and therefore adds value. Ants demonstrate this on a daily basis. In just a short space of time, they manage to find the optimum way from their nest to a place where food can be found. While searching for food and returning to the nest every ant secretes pheromones. So an ant leaving several scent marks on a short back and forth trail will mark the path more intensively than an ant can on its way to a remote source of food. The pheromones attract other ants to this trail and they’ll mark it as well. The faster this cluster of scent marks grows, the more ants will follow this short path to food. This starts a self-steering process. As the volume of individually generated information grows, it becomes increasingly meaningful and useful to the collective. An individual ant would not be able to achieve this.
Advantages of platooning
- The safety gaps between the trucks of a platoon can shrink from 50 (164) to ten meters (33 ft), cutting the space required by a convoy of three trucks in half.
- Slipstream driving in a platoon can improve fuel economy by eleven percent.
- Due to the “concertina effect” during braking and acceleration, experts regard ten trucks as the upper limit in a platoon.
- In the event that 50 percent of a truck’s annual mileage of 150,000 km (93,206 mi) should be driven in a convoy, every one of the linked trucks could save about 2,000 liters (528.34 gal) of diesel per year.
- About 90 percent of all truck accidents are partially or fully based on human failure. Platooning could clearly reduce this rate.
We can experience a similar process while driving our cars. Information about traffic jams is frequently based on position reports transmitted by smartphones. Every single one of them communicates that it’s not moving at the moment. And if a large number of them consecutively indicate “I’m not moving” the exact location and length of a traffic jam can be calculated quickly. This data pertaining to immobility benefits the mobility of the collective – a sufficiently large amount of information creating a new context. None of these phones reveals anything else but its location and currently immobile condition. When this information is combined, it creates a reflection of the traffic situation to which others can intelligently respond. While no individual is able to generate this information as value-added, every one of them is necessarily involved in creating it.
30 legs and 42 wheels in a platoon
When ants migrate a special kind of behavior can be observed – one from which a lot can be learned for our mobility. There are two conspicuous phenomena: Ants always migrate in small groups of five or six individuals that maintain a gap between their own and the preceding group, but will never pass it. This guarantees largely uninterrupted movement even in the event of short-term stagnation. Plus, in the event of cramped conditions the animals will directly contact each other using their feelers which enables them to keep moving forward at a uniform pace even in tight spaces.
As early as in 2012, Volvo, in the “Sartre” project (Safe Road Trains for the Environment), tested how this might be applied to road traffic by having several mutually coordinated vehicles travel in a convoy. And in 2016, all European commercial vehicle manufacturers demonstrated in the Platooning Challenge that the operation of road trains interlinked by means of electronic data transmission can work. Developers of smart vehicle control systems such as Highway Pilot Connect have transferred the behavior of ants in the form of platooning, i. e. moving in small groups, to road traffic, with several vehicles lining up behind a lead vehicle. Connected to each other via WLAN, almost like by an electronic drawbar, the assistance systems of the first vehicle will pass on all the key information to the following vehicles.
Less space, less consumption
Driving in such convoys – three semitrailers are typically used in tests – yields several advantages, as Thomas Grimm, Vice President Product Management Heavy Duty at Schaeffler, explains: “It reduces the aerodynamic drag of the vehicles following in the slipstream, enhances safety and reduces the required road space.” In fact, platooning can reduce CO2 emissions and fuel consumption by as much as eleven percent. Instead of the normal 150 meters (492 ft), a convoy is only about 80 meters (262 ft) long. This uses the available road space more effectively and the vehicles move at a very uniform pace. The lead vehicle issues a signal when interventions by the drivers of the following vehicles are necessary. This enhances safety, as a driver’s reaction time is 1.2 seconds whereas electronic systems provide information to the assistance systems of the following vehicles within 0.1 seconds. But even the best response times require braking systems that ensure reliable deceleration – especially in the light of the minimal gaps in a platoon convoy. “Schaeffler supplies important components for the reliability of brakes,” says Grimm. “The detection of axle loads that serves to issue warnings of improper cargo loading which may lead to longer braking distances and critical braking performance in emergency situations are topics we’re working on at Schaeffler as well.”
Swarm intelligence on the road
Platooning is a form of applied swarm intelligence – yielding benefits such as reduced fuel consumption, better traffic flow and higher safety. All the technologies required for field operation are available and, to some extent, already installed in the vehicles. “It’s conceivable,” explains Grimm, “that in the near future smart trucks will autonomously coordinate each other, forming platoons for certain routes, without requiring higher-level coordination.” The environmental benefits, improved road utilization and optimized driving strategy based on a wide set of data speak for the introduction of such systems.
However, in many places, traffic regulations still oppose a wider deployment of platooning systems, frequently stipulating a minimum distance between trucks that is clearly above the 15-meter (49-ft) gaps between the vehicles in a platoon. However, in view of the major advantages of platooning, there should be little doubt about forthcoming adjustments of legal provisions and in that case, truck platoons on our highways will be as natural as the small ant platoons have always been on forest soil.
Swarm intelligence at Schaeffler
“In Schaeffler’s view, one of the keys to forward-thinking concepts in road traffic lies in swarm intelligence,” says Thomas Grimm, Vice President Product Management Heavy Duty at Schaeffler. “Hundreds of thousands of data-producing commercial vehicles are constantly traveling our roads. We, as a component manufacturer, as well as freight haulers and vehicle manufacturers, regard the collection and analysis of this data as a mission in order to derive maximum benefit for the environment, the road users and the fleet operators from this data.”
Grimm uses four examples to illustrate the point:
Safety: Critical road surface or weather conditions (ice, aquaplaning, heavy rain, fog) can be detected by sensors and and passed on to the following traffic via a data cloud. The same applies to changes in traffic flow (formation of a traffic jam, end of a traffic jam). Grimm: “Anyone who has ever had a practice unit in a truck and experienced how long the braking distances can be and how helpless a driver is in such a situation understands the importance of this issue.”
Driving strategy and shifting behavior: When vehicles use smart data collection and coordinate their driving strategies, the environmental burden can be reduced. “Every additional shifting event under full load on steep inclines costs about a liter (0.26 gal) of fuel,” explains Grimm. “And every unnecessary braking event costs unnecessary energy as well. The more predictively and uniformly a vehicle moves the more efficient and less harmful to the environment it is.”
Maintenance intervals: “Sensor data and driving data can be gathered by vehicles and made available to all users,” explains Grimm. “The safety windows of wear parts could be optimized and adjusted to the respective vehicle. Components would no longer be exchanged according to time and mileage based intervals but only at the end of their actual useful life. This can provide operators of large fleets with major savings and enhance safety.
Road maintenance: Via the vehicles’ onboard sensors, e. g. in rolling bearings, road damage could be detected and repaired at an early stage. This can help avoid major road restorations that are costly and time-consuming.
In a nutshell: Vehicles are becoming increasingly smart and new markets are developing. “Schaeffler contributes its longstanding experience as a supplier there and combines it with new technologies of digitization,” says Grimm. “This also includes future business models such as the gathering, editing and provision of data to partners.”