Real-Time AI, your key to fleet management success
Machine Learning with the use of Artificial Intelligence (AI) could be explained as a system or machine which mimics human intelligence by performing iterative tasks which are continuously improved based on the learnings of collected information or data.
Although Hollywood movies have frequently portrayed AI as the demise of humanity (as human emotion is added to human intelligence), this thriving new technology is not meant to make humans obsolete, but rather to optimize human contributions to its highest level by recognizing patterns in data to enable predictions.
And if this is implemented into your company the correct way, it can be a very valuable asset for your business, and this certainly includes the world of fleet management. It can save your company money by optimizing safe driving and reducing accidents, managing your fleet reserves, cutting down on spending, and much more.
Digital transformation, however, has been fighting an uphill battle, according to Accenture. The necessity to monitor and evaluate the ever increasing amount of data has outstripped the capacity of a human brain while demanding exorbitant time and resources, the global consultancy firm said in a 2020 study "Human Impact of Data Literacy," and the situation remains quite the same today.
Currently, vehicle data is being collected and displayed on dashboards properly, but the problem is that companies do not have the capacity to monitor the data in real-time, and therefore are missing on the opportunity to optimize their business, according to computer scientist Anatoly Volkhover (pictured above).
In terms of fleet, some of the most common hidden inefficiencies nowadays are associated with the use of fuel, the utilization rate of equipment, and improper use of vehicles which leads to more frequent service calls, increased tire wear emissions, and reduced longevity of equipment.
“Another common inefficiency which can lead to wasted spending is a lack of predictability owing to bad weather or other disruptive incidents. All of this can be planned around with the proper tools,” says Mr. Volkhover who is CTO of business intelligence platform Prophecta.
To overcome these barriers, AI in the mainstream form (neural networks) or to say statistical learning will not help. What is needed is a highly specialized AI based on other principles which involve casual inference.
As for Prophecta, it offers real-time AI services, one of which involves creating a Digital Twin of the business or to say a causality-enabled mathematical model of their fleet operations. It is not that different from the mathematical models NASA creates for spacecraft except it is for vehicle fleets.
“It allows us to supply predictive analytics to fleets, and to apply corrective actions in real time to save money as well as preserve the environment,” Mr. Volkhover told Global Fleet all while explaining a recent Prophecta impact Case Study on a large energy company.
Case Study: Fortune 500 Oil & Gas Customer
Annual Fuel Budget
47% ($7,500,000 annually)
CO2 Emission Reduction
29% (2,000,000kg annually)
fewer traffic accidents, reduced personnel churn, extended equipment lifespan
In this case, much of the operational inefficiency was due to the need to cut down on idling by turning engines off and the need to better manage in-cabin heater usage. However, remember that the operational profile in every fleet is different.
A look at technology, in the eyes of Prophecta
• 5G is helpful but it is not required
• Cloud computing makes AI far less expensive. Use "serverless computing" by leveraging the huge capacity of the Cloud on a pay-per-use basis.
• As opposed to computer vision, process video streams. It benefits clients more profoundly if done right.
Data Privacy, a concern that can be overcome
Considering the powerful data driven tools used by AI, we cannot go without addressing privacy and security issues associated with connected vehicles and the need to protect the information of drivers and surrounding parties.
One company doing just this is Privacy4Cars, the first technology company to be highly focused on addressing data privacy. In its latest patent granted by the U.S. Patent & Trademark Office, the US-based company states the need to remove private information from a vehicle through a user computing device.
As for Prophecta, it is built in a way that the customer's data is owned by the customer and never leaves the customer's ownership.
“Our AI doesn't need to know the real identity of the drivers and other users. It works with the Digital Twin which only has a virtual equivalent of all participants,” says Mr. Volkhover, adding that no real names or personally identifiable information is used.
The information never leaves the customer's possession and is only merged into reports and dashboards when requested. Although Prophecta is based out of the US, it has recently been expanding into other regions such as Latin America, Europe, and more.
Today and Tomorrow
The elements that will drive telematics in the coming ten years will be Control, Compliance and Consent, according to Dr Andrew Jackson (pictured left) who is Research Director at the Ptolemus Consulting Group.
Control is crucial for ensuring good driver behaviour, which leads to the safety of people and goods being carried. And according to a recent survey by Arval Mobility, fleet managers can save up to €10,000 ($10,000) per fleet vehicle by detecting KPIs and taking action.
Among the KPIs to keep track of are those associated with dealing with severe weather, fuel costs, staff turnover, vehicle utilization, driver compliance, excessive idling, accelerating and braking, and safetey concerns, according to Mr. Volkhover.
So, fleet managers, develop your KPIs and take advantage of technologies such as AI so that you can start saving your company more money and be better prepared for the future.
To learn more about essential tools for Global Fleet success, register now and attend Global Fleet Conference VEx 2022 on 8 December for free. See you there!