The Drivaware platform can be used for many purposes, such as knowledge building, data geo-referencing, fleet management and ordinary and extraordinary maintenance.

Edit

Knowledge Building

The statistical analysis of telemetry data is useful, for instance, for getting information about fuel consumption and pollutants emissions. The Drivaware platform is able to acquire and store large amounts of data for long-lasting data acquisition campaigns on large fleets of vehicles.

The distribution of vehicle speed versus the engaged gear can characterize the driver's behavior as it is shown in Figure 1 where, for three identical cars driven by different users, the speed distribution is displayed jointly with the engaged gear distribution within each speed range. It clearly appears that User 3 preferably uses low gear ratios and thus high engine speeds, while — on the contrary — User 1 uses higher gear ratios at the same vehicle speeds. User 2 is similar to User 3.

Figure 1 - Speed/Gear Distribution

Figure 2 - BMEP Profiles

A data acquisition campaign on another set of three identical SI 2 liters cars, during around 250,000 km, allowed to highlight differences on actual engine load. In Figure 2 the time distribution of low (0-4 bar), medium (4-8 bar) and high (>8 bar) BMEPs — Brake Mean Effective Pressure, i.e. load — is shown for two different users in the set. User B requires higher loads compared to User C and thus can benefit from higher engine efficiency.

Other applications in user characterization are show and described in the figures below.

Figure 3 - Mean Time Between Engine Power-on

Figure 3 shows the mean time between engine power-ons. As one can see, the chart shows how frequently the car is started. Since the delay between two subsequent starts is greater the 4 hours in average, each duty cycle begins in engine cold conditions, having a significant impact on tailpipe pollutant emissions.

Figure 4 - Clutch and Brakes Operations

Figure 4 gives information on clutch and brake pedal usage in order to estimate wear of related mechanical components. For instance User 3 mostly drives in urban areas thus stressing both brakes and clutch more than User 2 who is mainly performing extra-urban trips. The same approach can be used for analyzing and predicting the wear of different engine and vehicle components such as gearbox, HVAC system, headlights and taillights lamps.

Figure 5 - Misc. Statistics

The first chart in Figure 5 gives the number of starter motor activations every 1000 km and allows to forecast its wear and to plan maintenance interventions. The second and third charts highlight the reverse gear usage and the amount of time the clutch pedal was held pushed respectively. It could be interesting to notice that, on average, 20% of the total usage time the engine is disconnected from the wheels.

Figure 6 - Characterization of Route Types

Much more advanced analyses can be performed, as shown in Figure 6 which reports the speed vs. acceleration distributions on several vehicle duty cycles (red refers to most frequent conditions, while blue to least frequent ones). Records 8, 11 and 14 (A) were generated in congested urban conditions in Milan (Italy) while the others (B) also contain a highway part. Similarities among trips are clearly visible.

Extensive tests were performed on fuzzy filtering of previous distributions so that relevant characteristics could be highlighted. At the same time pseudo-hierarchical data clustering was applied after having defined a suitable distance between distributions.

Edit

Geo-referencing

The Drivaware platform has built-in GPS capabilities; it's therefore possible to reference vehicle data to geographical positions.

Geo-referenced data contributes widening the functionalities and possibilities offered by the platform. It's even possible to track vehicle usage and monitor routes as they are driven, in a quasi-real-time fashion.

In the picture on the right, the points with speed 0 km/h have been highlighted in black in the vehicle route, clearly displaying possible congestion points in the city's road network.
Edit

Fleet Management

One of the possible applications of the Drivaware platform is fleet management. In this case, the platform allows to reduce costs and improve efficiency.


Edit

Ordinary Maintenance

The most important goal of ordinary maintenance is to accomplish to all the time- and distance-based tasks such as oil change, filters replacement, tires check-up, etc.

Vehicle warranties are usually valid if all the maintenance tasks are performed regularly. The Drivaware system can read the odometer and therefore generate and notify events at predefined mileage values. Applying this approach enables to automatically manage maintenance tasks planning and organization, avoiding infrastructure congestion and notifying users and staff on time, without manual intervention.

Edit

Extraordinary Maintenance

Taking care of extraordinary maintenance tasks, such as faults and anomalies, is very important: faults represent high costs for the fleet manager and also reduce QoS.

Drivaware monitors the OBD bus for fault signals (DTCs), reporting and notifying such situations to the fleet administrators. The early detection of anomalies and faults effectively cuts costs especially by reducing the MTTR (Mean Time To Repair) and by avoiding severe faults caused by unfixed minor issues.