roko

Getting past traffic woes

The Microsoft Research India’s Nericell project tries to collect information about not just speed of the traffic flow in a particular location but also how noisy it is, or how bumpy the roads are, helping you to choose not the fastest but the safest and most convenient route.

Traffic woes are a fact of life in all major Indian cities. Dynamic road infrastructure worsens things, with people being clueless about new roads/ flyover etc built every now and then. Taking a cue from the western world to solve our traffic problems may not be wise owing to varied road conditions (e.g., potholed roads), chaotic traffic (e.g., a lot of braking and honking), and a heterogeneous mix of vehicles (2-wheelers, 3-wheelers, cars, buses, etc.).

When a bunch of researchers from Microsoft landed up in Bangalore from the US, the first thing that came to their notice was, of course, the traffic. They then decided to create a low-cost and efficient solution to remove traffic snarls from cities like Bangalore. The first thing that struck them was the prowess of the s that are available. Smarts have not only become more of a computer but also great sensing devices, with things like a noise detector (the micro), GPS, and accelerometer. Although the penetration of such s in India is low at present, the growth is phenomenal. Considering the capability of s to send out information about their environment using these sensors, the researchers began working on a project called Nericell.

Understanding Nericell

Nericell addresses several challenges including virtually reorienting the accelerometer on a that is at an arbitrary orientation, and performing honk detection and localization in an energy efficient manner. Prior work in this area has primarily focused on the developed world, where good roads and orderly traffic mean that you can largely characterize traffic conditions on a stretch of road by the volume and speed of traffic flowing through it. To monitor this information, researchers have developed intelligent transportation systems (ITS). Many of these involve deploying dedicated sensors in vehicles (e.g., GPS-based tracking units) and/or on roads (inductive loop vehicle detectors, traffic cameras, Doppler radar, etc.), which can be an expensive proposition and so is typically restricted to the busiest stretches of road.

To address this challenge, Microsoft is working on Nericell, a system for rich monitoring of road and traffic conditions that piggybacks on smarts. Explaining how Nericell works, Venkat Padmanabhan, Senior Researcher Microsoft Research India, said, “Nericell orchestrates the smart to perform sensing and report back to a server for data aggregation. Smarts include a range of sensing and communication capabilities, in addition to computing. A might include any or all of a micro, camera, GPS, and accelerometer, each of which a developer can use for traffic sensing functions. In addition, the includes a cellular radio (e.g., GSM), possibly with data communication capabilities (e.g., GPRS or UMTS).”

A -based approach to traffic monitoring is a good match for developing regions because it avoids the need for expensive and specialized traffic monitoring infrastructure. It also avoids dependence on advanced vehicle features such the Controller Area Network (CAN) bus that are absent in the low-cost vehicles that are commonplace in developing regions (e.g., the 3-wheeled autorickshaws in India). Moreover, it takes advantage of the booming growth of telephony in these regions. For example, India has about 300 million subscribers, growing at an estimated 7 million per month. Although the majority of these users have basic s today, a large number of them, in fact more than the number of PC Internet users in India, access the Internet on their s. [Smart sales are currently running at about a million per quarter and this is likely to go up as touchscreen s become popular and manufacturers introduce entry-level smarts in the Rs 12,000-15,000 price bracket. – Editor] There are similar growth trends in many other parts of the world, with the total number of subscriptions worldwide estimated at 3.3 billion.

“Despite being -based, our approach to traffic monitoring is distinct from prior work based on remote tracking of s by cellular operators. Our sensing and inferences goes beyond just monitoring location and speed information, hence requiring a presence on the device,” Padmanabhan clarified.

Several technical challenges arise from this design choice to perform rich sensing and base the system on smarts. Nericell leverages sensors besides GPS—accelerometer and micro, in particular—to glean rich information, e.g., the quality of the road or the noisiness of traffic. The use of an accelerometer introduces the challenge of virtually reorienting it to compensate for the arbitrary orientation of a smart. Furthermore, we need to design efficient and robust bump, brake and honk detectors in order to infer road and traffic conditions.

“Moreover, since a smart is battery powered and is primarily someone’s , energy-efficiency is a key consideration in Nericell. To this end, we employ the concept of triggered sensing, wherein we use a sensor that is relatively inexpensive from an energy viewpoint (e.g., cellular radio or accelerometer) to trigger the operation of a more expensive sensor (e.g., GPS or micro). For efficiency in communication and energy usage, each node processes the sensed data locally before shipping the processed data back to the server,” Padmanabhan said.

The main contributions of this work are:

Algorithms to virtually reorient a disoriented accelerometer along a canonical set of axes that then use simple threshold-based heuristics to detect bumps and potholes, and braking;
Heuristics to identify honking by using audio samples sensed via the micro;
Evaluation of the use of cellular tower information in dense deployments in developing countries to perform energy-efficient localization; and triggered sensing techniques, wherein a low energy sensor is used to trigger the operation of a high-energy sensor.

Related work

Intelligent transportation systems have been proposed and built to leverage computing and communication technology for various purposes: traffic management, routing planning, safety of vehicles and roadways, emergency services, etc. We focus here on work that is most relevant to Nericell.

There has been much work on systems for traffic monitoring, both in the research world and in the commercial space. Many of these systems leverage vehicle-based GPS units (e.g., as in GM’s OnStar) that track the movement of vehicles and report this information back to a server for aggregation and analysis. For instance, CarTel includes a special box installed in vehicles to monitor their movements using GPS and report it back using opportunistic communication across a range of radios (Wi-Fi, Bluetooth, cellular). Applications such as route planning then use this information.

Recent work on Surface Street Traffic Estimation also uses GPS-derived location traces but goes beyond just estimating speed to identifying anomalous traffic situations using both the temporal and spatial distributions of speed. For instance, the authors are able to distinguish between traffic congestion and vehicles halting at a traffic signal.

Operational services, both commercial and otherwise, have been built using GPS information as well as information from other traffic sensors deployed in an area (inductive loop vehicle detectors, traffic cameras, Doppler radar, etc.). Examples include the Washington State SmartTrek system in the US (which includes, among other things, the Busview service, to track the city buses in the Seattle area), and the INRIX system for predicting traffic based on historical data.

There has also been work on leveraging s carried by users as traffic probes. Smith et al report on a trial conducted in Virginia in 2000, which was based on localizing s using information gathered at the cellular towers. This study made a number of interesting observations, including that the sample density (at the place and time of this study) was only sufficient for estimating speed with moderate accuracy (within 10 mph = 16 kmph), and that there was a tendency to underestimate speed because of samples from stationary s located near the roadway. Regardless, the rapid growth in penetration has spurred the deployment of similar systems in other locations worldwide, including in Bangalore.

Much of the work on ITS has used GPS-based localization and some of it has used localization performed at cell towers.

A unique approach

“Our work on Nericell builds on prior work but is distinct from it in several ways. We do not replicate the significant body of prior work on estimating the speed of traffic flow and driving patterns based on location traces, and presenting this information to users in an appropriate form. Instead, Nericell focuses on novel aspects of sensing varied road and traffic conditions, such as bumpy roads and noisy traffic. Furthermore, Nericell uses smarts that users happen to carry with them, depending only on capabilities that are already available in some s and that are likely to be available in many more in the coming years,” Padmanabhan said.

By piggybacking on an existing platform, Nericell avoids the need for specialized and potentially expensive monitoring equipment to be installed, whether on vehicles as in or as part of the infrastructure as in SmartTrek. However, building on top of a platform introduces challenges, for instance, with regard to accelerometer orientation, energy efficiency and device localization, which the researchers have addressed in Nericell. Finally, Nericell falls under the active area of research called opportunistic or participatory sensing and can leverage ongoing research on challenges that are inherent to this area, e.g., data credibility and privacy.

The richness of sensing that Nericell encompasses is motivated by the wide applicability envisaged for the system. You can use the system to annotate traditional traffic maps with information such as the bumpiness of roads, and the noisiness and level of chaos in traffic, for the benefit of the traffic police, the road works department, and ordinary users. For instance, a user might search for a route that minimizes the number of chaotic intersections to be traversed, thereby optimizing for ‘blood pressure’ rather than for distance or time. While these applications serve as the motivation, the focus of the project is on the sensing component of Nericell, specifically, on how to efficiently use the accelerometer, micro, GSM, and GPS sensors in smart s to detect bumps and potholes, braking, and honking, and to determine location in an energy-efficient manner.

Padmanabhan opined, “Given the energy costs of the different sensors, Nericell only keeps its GSM radio (which has to be turned on anyway for the to function) and the accelerometer turned on continually. It uses input from these two devices to trigger the turning on of the other sensors, for instance, to obtain a precise location fix using GPS. Note that each individual filters and processes its sensed data locally before reporting it to a server for aggregation. This not only cuts down on the cost of communication (including the energy cost entailed), it helps with privacy (for example, by not uploading raw audio feeds) and it cuts down on the volume of extraneous data, not reflective of road and traffic conditions that the aggregator has to contend with. For example, we filter out accelerometer readings arising from a user fidgeting with their rather than from bumps in the road.”

Overcoming challenges

One of the biggest challenges for the Nericell project has been to contain the battery drainage of the s. To resolve this problem, Nericell using a triggered-sensing technique, wherein a low-energy sensor can trigger a high-energy sensor. These sensors can be of difference type. For e.g. GPS is a battery intensive sensor; to track a vehicle in a smooth traffic, accelerometer, which is a cheaper sensor in terms of battery life, is used to check to bumps. When there are too many bumps, GPS can be triggered to get the actual location of the vehicle. Similarly, the cellular tower radio can also be used to track the location of a vehicle, unless the GPS is required.

Another challenge, towards which the team is constantly working, is to remove noise from the system. Now, the Nericell system reports a lot of events or false positives, which need intelligent filtering. For e.g. In case the drops, it will be difficult to understand whether it was a drop or was it a bump.

Going ahead, Nericell will focus on the use of camera to provide real-time picture of the traffic situation in a particular place. However, this process would require human intervention and cannot be automated, as the camera needs to be manually pointed towards the subject. Microsoft Research lab India is also trying to improving the sensing capabilities of the technology to eliminate any room for error.

varun.aggarwalexpressindia.com