Artificial Intelligence. Machine Vision. Big Data. There’s a lot of buzz around advanced technology being applied in nearly every sector from construction and field services, to trucking and logistics. Google’s CEO Sundar Pichai recently said the development of AI was more profound for humans than fire and electricity.
But let’s talk about the bottom line: how can these technologies, when used together, create opportunities to most effectively drive down costs and generate revenue?
Going beyond video
The most advanced machine vision algorithms can be trained to detect just about anything the human eye can see. Fleets have used this capability to optimize safety training and driver retention, and also collect evidence that can help lower liability costs in traffic incidents and reduce insurance premiums.
But this technology excels when data, machine vision (MV), and artificial intelligence (AI) are used in concert (MV+AI) to provide a more holistic context to events. As an example, a collision happened when a fleet vehicle collided with a car while both vehicles were merging onto the freeway, making it difficult to establish the sequence of events just based on video.
When combining the metadata from the video feed with the ECM data from the vehicle, and predictive data from a MV stream, the usage of turn signals and acceleration/deceleration by the driver were related to the event.
This data, combined with the video, helped establish that the carrier’s driver did not cause of the collision. When we solve patterns like this, we are able to further use machine learning and artificial intelligence to extend this sort of pattern analysis to other similar, unrelated events at large scale to automatically establish their contexts.
Identifying customer needs
MV+AI can also detect otherwise innocuous elements in an environment that, when combined with customer data, reveal opportunities for fleets to expand services to existing clients or even generate new customer leads.
These can arise not only during a driver’s route but at the point of dispatch or delivery. A waste disposal company that uses MV+AI could detect when a customer’s bins are overflowing and automatically suggest additional collection service that the disposal company can provide within the customer’s budget.
The company could also detect overflowing bins at locations along its regular routes and, combined with mapping data, reveal potential new customer leads or upsell opportunities. As an example, intelligent video can help identify specific areas where overgrown trees need to be trimmed away from power lines. Another example, related to municipal services and safety, is detecting broken traffic lights, faded lane markings, or missing traffic signs.
Ever since the invention of the fuel gauge, monitoring the mechanics of vehicles has been a key part of optimizing vehicle management and reducing costs.
But now we can go a step further. Pairing data from vehicle monitoring systems connected to the engine control module with a predictive AI means fleets can anticipate and prepare for maintenance. Optimizing maintenance operations also helps to prevent risks such as an engine failing while operating at 60 mph on a freeway, which can create a dangerous situation for all drivers on the road.
The further ahead fleets can plan repairs, the better position they are in to minimize the costs and disruption to operations. And the less time a service vehicle is in the shop, the more time it can spend in the field serving clients and generating revenue.
Reducing internal silos
Using all three – MV, AI, and data – together and concurrently across a fleet unlocks the true value of these technologies.
Having a single pane of control across company departments allows for the reduction in internal silos and reduces the risk that one priority for the company is optimized at the expense of others. In fact, using data in isolation can become detrimental.
An example that comes up time and again is the importance of compliance and safety departments using the same technology and data to make decisions in real-time. Drivers, for example, are limited in the number of consecutive hours they spend behind the wheel for compliance reasons. If fleets only optimized for compliance, a siloed system would just select the next available driver, rather than the next available driver with an appropriate safety rating.
Neglecting to look at complete data can be a major pitfall for companies and can result in leaving money on the table or increasing rather than reducing operational risks.
The future of data
Data is often the biggest asset a company has, and we anticipate that in the future more companies will use it as a source of revenue itself.
All operational companies, whether freight or otherwise, have something in their data that would be useful to other segments in their industry. The data on the maintenance costs of a 5,000-vehicle fleet could, for example, provide valuable lead generation insights to a maintenance company. The safety data of drivers in different geographic areas would allow insurance companies to more accurately assess the risk scores of potential clients in the region.
In the future, smart cities will include this type of data, too. A fleet’s route data on hazards or traffic would be crucial for local governments when it comes to city planning or providing services such as waste or tree removal. Fleets might eventually be buying anonymized data from other fleets as a way to share intelligence on equipment that both use and uncover patterns that help optimize their use and lifespan.
Provided that the proper steps are taken to de-identify data, the future points to more robust data marketplaces where companies and even fleets turn their data into a valuable revenue generator.
As MV+AI and data capabilities become more commonplace and widespread, the fleets that can use it most efficiently to capture revenue and break down silos will have the edge.