Battery Analytics Proves a Needed Tool for EV Fleets

Although much of the news surrounding EVs this year has been around a slowdown in sales, with many manufacturers scaling back their production plans, those headlines crowd out what should be the main takeaway: Long-term growth forecasts are still solid, and the EV revolution is well underway.

Each year continues to see record EV sales. Even without significant policy changes to tighten emissions restrictions, most major markets are set to see substantial expansion of EV uptake in the next decade or so.

By 2035, Europe should surpass China in EV adoption, with over 80% of passenger car sales electric. Even in the U.S., where limited charging infrastructure and range anxiety have stunted EV growth to some degree, 60% of new cars by 2035 will be EVs.

The continued growth of EVs also applies to delivery, passenger transportation, and heavy-duty vehicles, as advancing battery technology means that electric heavy vehicles are increasingly competitive with diesel engines in terms of cost and efficiency. By 2035, electric buses will make up most bus sales in the three major EV markets, with electric trucks accounting for half in the U.S. and China. Even in emerging EV markets, electric buses will dramatically outstrip uptake for passenger vehicles.

This speaks to how fleet assets like city buses are especially well-suited to electrification, with fixed driving patterns, shorter driving distances compared to other heavy vehicles, and aggressive municipal emissions goals keeping interest in electric buses high.

Outside of buses, even the heavy vehicle segment many expected to be the hardest to electrify — Class 8 trucking — will soon be competitive with diesel-powered trucks.

Fleet Growth Raises Safety Concerns

The rapid expansion of EVs also brings heightened focus on safety and reliability due to concerns around lithium batteries. This is especially true for stakeholders operating urban or passenger EV fleets, where dozens of riders per bus and/or dense city environments increase the level of risk.

Battery faults that take a vehicle out of operation can also disrupt deliveries, commutes for thousands of people, or cause financial damage. Even if an issue is not as severe as a fire, an unreliable EV in the fleet can still cause great reputational or financial damage to the operator.

Although EVs are far less likely to experience fires than internal combustion vehicles, fires are an ongoing concern with lithium batteries. The prolonged nature of lithium battery fires can create self-sustaining conflagrations. This may require distinct firefighting and has resulted in several high-profile incidents. The media may overstate the risk of battery fires, but it’s a unique one that any operator of EVs needs to be prepared for.

As a result of reliability concerns, there is a much incentive for EV fleet operators to do all they can to avoid a fire or a battery issue that requires idling their fleets. Comprehensive battery analytics software is the key to operating and managing EV batteries in a way that maximizes their lifespan and minimizes fire and failure risks.

Complex Batteries Require Advanced Software

EVs appear to be much simpler vehicles than diesel ones, but the reality is that a complex web of hardware and software is enabling the lithium battery to power the vehicle.

The devices, cables, sensors, and cooling systems that connect the battery with the rest of the vehicle are just as critical as the battery, itself a complex web of cells – and all can potentially hurt the battery. Ideal EV operation also assumes perfect charging infrastructure, which is not always the case, and issues like overcharging can eventually lead to a safety fault. This is to say nothing of the battery itself, which for fleet users could come from multiple manufacturers, all with individual quality assurance processes.

Advanced battery analytics platforms account for all these factors to continuously monitor and analyze issues wherever they may appear. Compared to a traditional battery management system (BMS), which manages only a limited number of risk factors and is a reactive tool, analytics enable fleet owners to practice predictive maintenance.

By drawing from a wealth of battery research and combining it with artificial intelligence algorithms, analytics software builds a profile of each battery’s state and forecasts its future behavior based on current and past conditions.

Fleet owners can see problems in advance and adjust operations to accommodate vehicle maintenance, rather than reacting to issues after the fact that unexpectedly take vehicles out of duty. The insights offered by analytics further enable fleet operators to anticipate downtime by identifying issues that could lead to fires or other safety faults, providing the tools and spare parts needed to repair a bus or truck battery before an incident.

Battery Management Software Limits

While all lithium batteries in EVs include a BMS that offers some battery monitoring, fleet owners should be wary of relying on this alone to prevent safety or reliability issues. Fleet managers should understand two key distinctions:

1. The BMS is designed to run a single EV; battery analytics software is there to run the fleet. BMS software can fail or perform poorly, leading the battery to function in a manner that eventually leads to problems before the BMS detects them. Manufacturer-provided systems are also limited in what they can monitor; for example, they cannot detect or prevent cell reactions that might lead to a fire-inducing short circuit. BMS software also provides limited useful data to fleet owners, instead focusing on simply ensuring the battery is operating. Their inability to align battery operations with long-term predictions based on past and current usage makes it dangerous to rely on a BMS alone to ensure that a fleet owner’s changing operational plans do not lead to a safety incident.
2. A BMS is also rather simple in its design, which is less than ideal for scenarios that may have once been edge cases but are becoming more commonplace. As the technology of EV batteries and charging infrastructure advances, fast charging is now increasingly common. However, it can lead to faster battery degradation. A BMS cannot adapt its charging protocols as the battery degrades. Equipped with only a BMS, batteries are at a higher risk of unhealthy charging practices, potentially leading to safety risks like short circuits.

In complex cases, battery analytics software complements the BMS and provides a much more comprehensive toolset for avoiding safety and reliability issues.

Fleet Safety and Reliability Depend on Analytics

Installing battery analytics is an ideal way to ensure that fleet operators understand the complex causes of battery fires and failures, and that they can tackle them before they cause catastrophic damage to an EV fleet.

Compared to a BMS, analytics provide continuous monitoring that considers expert battery research and the historical performance of the battery. The monitoring routines use this information in an AI algorithm to identify anomalous trends that could lead to a fire or failure, and immediately alert the battery operator. The analytics suite presents this information to the user in an easy-to-understand dashboard that collates alerts by severity so the most serious issues can be addressed immediately.

The predictive maintenance enabled by battery analytics can help fleet managers keep vehicles safe and with minimal downtime. An electric vehicle equipped only with a BMS may avoid a fire when the software shuts down the battery due to a fault, but then the vehicle is unexpectedly out of service, forcing the fleet manager to suddenly adapt routing with fewer vehicles. If the BMS fails to detect the issue, the resulting fire could be disastrous for the operation. With battery analytics, the software identifies the issue much earlier, and the fleet managers can adjust operations well in advance to bring the vehicle in for repairs.

Two Examples of Fire Prevention

1. In one example, a company’s analytics software was able to review data from a battery fire and determine that a fire was likely to occur six months before the actual incident. The analytics ultimately determined the cause to be a manufacturing defect that quality assurance at the factory and the battery’s BMS missed. Even beyond a dramatic safety incident, analytics can also help with reliability issues that are just as crucial for fleet operations.
2. A bus operator recently deployed analytics to try and determine why the charge level of its electric bus seemed to be off, with unusually fast changes. The issue resulted in the battery halting its charging early, as the state-of-charge signal mistakenly indicated the battery was full — a condition that’s problematic for a fleet operator, as the bus cannot complete its route without knowing the actual range. The analytics company found that the fault lay with an issue in the BMS firmware, which the fleet operator would not have discovered using the BMS alone.

Battery Analytics Maxes Uptime

EVs in general demand particular attention to battery safety and reliability, and this is especially true for fleets. Any fleet requires high uptime, and operators must predict when vehicles will be down for maintenance and how to avoid catastrophic fires. Advanced battery analytics provide fleet operators with this ability and are just as critical as the charging infrastructure required to start an electric fleet.