AI and related platforms such as ChatGPT appear to be a natural fit for electric vehicles, which generate and rely heavily on data.
The emerging AI technology could create more neural techno-pathways that streamline businesses and complement fleet operations with speedier solutions.
AI Advancing on Many Fleet Fronts
AI offers much potential in the maintenance area of fleet operations, as electric vehicles get smarter and create opportunities to absorb and aggregate data for new insights, said Chris Knosp, director of fleet electrification for Mike Albert Fleet Solutions.
“AI is bringing innovation forward in maintenance, routing and duty cycles,” Knosp said. “The metrics are changing as vehicles run longer and battery electric vehicles can lengthen duty cycles.”
At Mike Albert, the managers are still studying regenerative AI platforms before developing a strategy to use it in fleet operations.
Josh Green, CEO and founder of Inspiration Mobility, a company that works with fleets on electrification, said AI platforms for electric vehicle fleets should be taken seriously as they develop.
“We are already reinventing the fleet management company for the electric vehicle age, and part of that is the technology,” he said. “AI and Chat are the latest examples of how you can leverage the latest in tech and create nimble EV-only dashboards enabled by the data from EVs and chargers. There are many data streams beyond telematics.”
AI-Driven Fleet Electrification
AI could propel fleet electrification with its ability to compute predictive analytics, provide EV battery intelligence and monitoring, and map out charging infrastructure, said Steve Greenfield, CEO and founder of Automotive Ventures.
“It would be able to switch out electric vehicles on routes based on charging levels and availability, and smooth out the degradation of batteries,” Greenfield said. “There are all types of opportunities for AI to help fleet managers hit targets.”
Other AI uses becoming common in fleet operations include software and routing providers using regenerative technology to process data for TCO, find the best routing, and manage EV charging solutions, said Green of Inspiration Mobility.
Electric vehicles are more amenable to AI given how much data they produce. That can yield more powerful conclusions, lessons, and insights, he said.
“With an electric fleet, there are so many different data points. The EV itself gives off much more data, and now you have phone embedded apps, video telematics, the charger information, utility rate structures, and geofencing.”
As OEMs ramp up EV availability and production, they will add more data insights that require more numbers crunching and time horsepower, Green said.
“In that way AI and ChatGPT can accelerate things and provide more useful information. That is an area this should be more valuable.”
Fleet electrification overall is a “systems integration puzzle,” Green said. “It produces so much more robust data and room for error. With a wider array of factors, having more data with more complexity brings challenges.
"AI and ChatGPT can streamline those by processing more quickly and efficiently because they are based on natural learning models.”
ChatGPT Is Just the Start of the Revolution
Green outlined ways AI and ChatGPT could be used in electric and other types of fleet operations:
- Embed ChatGPT or regenerative AI into certain customer facing portals and apps to speed up response times.
- Capture more data and analysis that doesn’t require teams of human analysts.
- Enable fleet managers to see which EVs are best on certain routes and deployments.
- Update the TCO on fleet vehicles by prioritizing and tailoring TCO analysis on the latest data.
- Fleet managers can quickly get accurate forecast data and specs on the performance of new electric vehicle models coming out.
- Identify places and times to charge EVs and what level chargers to use for a particular EV’s duty cycle.
- Get real time alerts on how cold weather is affecting the range of different EV models in real time.
So far, Inspiration Mobility itself is using ChatGPT for writing marketing copy and creating internal documents, Green said. It’s also proven helpful starting and building a learning module for improved safety benefits of EVs.
“We have used GPT to produce a starting point on many questions and docs that will be saving people time. We’re building functionality into all customer facing apps.” Those would include a fleet manager portal, driver app, ESG portal — “all have some function and features that are driven by AI behind the scenes and enabled by ChatGPT.”
Greenfield sees AI being applied in sectors and in roles that would intersect with fleet operations, thereby spreading the intelligent design throughout the automotive sector:
- At the OEM level, AI could gather data from a production line in real time and identify flaws in the processes and immediately recommend or make corrections, which leads to big savings.
- Another application could anticipate leading signals of a possible failure or breakdown on a vehicle, which would help an operation save on down time. By collecting sensor data and building algorithms, AI could “tell you the timing to get a repair done proactively compared to what’s coming down the road,” he said.
- Anywhere anyone is doing knowledge repetitive tasks, the work should be complemented with interfaces like ChatGPT.
- ChatGPT and regenerative AI could also help refine vehicle inspections and reconditioning with corrective digital cameras, identifying vehicle damage and flaws in real time while providing to-do-to-repair checklists.
- Applied to robotics and automation, ChatGPT could speed up the performance and refine all repetitive tasks replacing manual work and rote white-collar work.
- Within business organizations, ChatGPT could take over some of the tasks of accountants, bookkeepers, and software developers. “Anyone pulling data in analytics departments, drivers, or managers in an organization, should feel enabled by ChatGPT.”
One of the main advantages of ChatGPT for fleets is how it can handle junior-level analysis, he said. It can save time compared to what a normal human can do by collecting data and delivering actionable insights.
Smartening Up Remarketing Logistics
AI already has been a leading tool in developing replacements for the manual inspections that have been common throughout the remarketing industry. That dovetails with the logistics of transporting vehicles among consignors, auctions, and dealers.
At Ship.Cars, a transportation management system for auto carriers that helps them find the best loads and routes, AI is used in an application that provides electronic proof of delivery and inspections to make sure the vehicles have not sustained damage. It also accurately tracks ETAs for individual vehicles. Ship.Cars transports about 50,000 vehicles per month, with 30,000 of them being inspected.
Vlad Kadurin, head of product at Ship.Cars, working out of Sofia, Bulgaria, has led all software product development and charts business strategy for the auto logistics digital platform, which ships vehicles for dealerships, auctions, and brokerages.
“When the driver goes to a pickup location, they do a thorough inspection and mark damage and track the exact date of the vehicle,” Kadurin said. “They do the same with delivery to prove the vehicle is in the same condition. They take 360 images and use AI damage inspection algorithm, to see damage on images itself.”
By eliminating the human factor, AI decreases the inspection time from 4-5 minutes to 1-2 minutes. Vehicle images are documented via 360-degree arc-arm cameras that capture all sides of the vehicle.
“Using this AI inspection with good images makes sure we find most of the damage,” Kadurin said. “From a statistical standpoint, there are fewer errors. The model can be tested on hundreds of thousands of images. There’s a less than 5% chance to miss the damage if visible.”
The main benefit of the AI-driven inspections is peace of mind among auto transporters, dealerships, and OEMs all wanting vehicles delivered in the best condition. Carriers can know whether a flaw was their fault.
“With the technology, you can get the condition of the vehicle without the human factor,” Kadurin said.
The AI systems also can ward off one of the biggest problems of false claims that hit insurers, brokers, and carriers. False claims can eat up time and costs to determine fault and mediate.
“It saves on immediate losses,” he said. “Even if an insurance company covers the damage, a holder’s insurance renewal premiums can go up. The shippers and brokers may stop working with a carrier or reduce volume and not consider them reliable, which hurts the reputation of the carrier.”
As to ChatGPT and similar regenerative AI platforms, Ship.Cars is evaluating the tool to see how it can enhance its services. One pursuit is a chatbot for carriers, drivers, and users to help inform the end-customer. The chatbot goes into databases and takes information and answers the questions like a human.
The primary challenge with AI now is to “eliminate the hallucinations,” Kadurin said, referring to the wrong results based on inaccurate information. “Most companies cannot use Chat GPT directly, but we can and should use ChatGPT logic in our help center for information on features and service. If you feed something to ChatGPT, it becomes usable to others.”
The success of regenerative AI depends on machine wording that creates human-like intelligence and behavior.
“We’re doing our own custom solution based on Chat GPT to provide a chat box human like experience,” he said, adding the company is also partnering with other tech players to develop the technology. “The purpose to eliminate first level support.”
Dealerships Save Time and Spare Reputations with AI
Auto dealers are ripe for AI applications that can reduce all the interactions and formalities of buying and selling vehicles.
One firm with early insights into AI and dealerships is Widewail, a reputation management and marketing trust company that works closely with dealerships to deploy technology that monitors, enhances, and organizes customer reviews and social media comments while recommending responses.
In one example, Widewail looks at the motives and reasons for positive and negative customer sentiments related to electric vehicles and their pricing.
“A local business’ most powerful asset is happy customers,” said Matt Murray, CEO of Widewail. “The challenge is how to harness that and activate their voice and advocate for the business. The voice of the dealership depends on how you manage public interaction, reviews, comments on social.”
AI can help see the trends in the reviews and assess customer reactions.
The company is understanding its capabilities, advising product and engineering teams to take a metered approach and be mindful of risk, Murray said. They are looking at ways ChatGPT can allow a customer data platform (CDP) to power the context of interactions with customers.
“With an automotive context to chatbot, it may be able to hold a useful and valuable conversation.”
Based on research of what customers value the most about a dealership, friendly and helpful interactions with staff rank the highest, Murray said. “You still need people involved with managing conversations. You can’t just turn on a bot and have great conversations. That is not the reality we live in.
“We’re holding public conversations with car shoppers and service customers on behalf of dealers,” he added. “The accuracy and value of response is very important because it is the customer experience.”
AI boosts productivity by suggesting responses to reviews and comments on Facebook, Google, and Yelp, said Adam Burnett, CTO of Widewail. But you still need human editing. “A thoughtful response to a two-sentence review is still a challenge and gets routine. If they see you using a bot, they’re disengaged from the customer experience and not getting a good impression.”
On the other hand, it’s easy for an owner or manager to get personal about a negative review, he said. “You don’t want to get into a war over a one-star review. A chat bot can protect dealers from themselves and prevent a catastrophe of a reply. A bot can get more info from users and direct them to a contact at the dealership.”
Murray said we are still accustomed to computers doing what humans tell them to do. Machine learning enables the computer to start showing creativity. “We’re starting to see why it does certain things.”
Murray warned that ChatGPT still contains risks, and it must be paired with human guidance. “GPT’s primary job is to satisfy your request. It will do it at the expense of accuracy. It will fabricate facts. It will lie to you. It lies almost 30% of the time.”
What it does well is what Murray describes as the “democratization of analytics.”
“You can feed GPT complex data and ask it to summarize a book. It can summarize the customer experience on a specific data set.”
ChatGPT is inexpensive in running analytics, Burnett said. “You can feed it as much tech as you want to generate summaries or topics that can apply to a particular industry.”
ChatGPT is still a long way from where chat bots can hold conversations with a human that makes it undetectable, Murray said. “What makes people most upset or happy is the quality of interactions in the sales or service process. I think we are at the very early stages of how technologies impact inventory and any place in automotive where we have structured data such as pricing.”
ChatGPT will prove valuable because e-commerce is hard to crack for automotive, Murray and Burnet agreed. The speed of the application will increase so that instead of comparing last year’s forecast to this year, you can do last week versus this week and next.
“I think these technologies will help corporations better prioritize their investments,” Murray added.
Originally posted on Automotive Fleet