From Insight to Impact: An Opinion on the Critical Role of Diagnostics in Li-ion Battery Production

Global Li-ion battery capacity is projected to reach 4 terawatt hours (TWh) by 2030, a fourfold increase from the 1 TWh milestone achieved last year.

As battery installations increase annually, production rates will rise to meet the growing market demand. This surge has prompted corporations to develop more gigafactories, which are GW-scale battery production facilities.

To support this rapid expansion, it is crucial to implement economical and high-yield manufacturing processes that produce high-performance Li-ion batteries with minimal waste, ensuring that production can keep pace with demand while maintaining quality and efficiency.

Battery waste generation is set to accompany increasing production rates

The rise in production rates inevitably leads to an increase in battery waste generation. New gigafactories, often producing specific battery cells or types for the first time, can have scrap rates as high as 70–90% during initial ramp-up due to defective battery construction, eventually achieving a 20–30% scrap rate during steady-state production. At 30% this equates to an annual loss of approximately US$225 million for a 10 GWh facility with recovery costs taking up to four years.

These defects can vary from misalignment of components inside the battery to defects in the electrode material such as wrinkles and pinholes to cross-web gradients in key electrode properties such as density, thickness and conductivity. In an industry characterised by razor-thin margins, such inefficiencies have a transformative impact on cost.

In some cases, defective batteries have reached the market, resulting in billions of dollars in losses for manufacturers, highlighting the urgent need for industry-wide optimisation. Notable incidents include LG Corp’s $2 billion recall of the Bolt EV and Hyundai’s $900 million recall affecting 82,000 electric cars.

The Big Picture

As Li-ion batteries become more sophisticated, so too must the methods to diagnose their properties and as battery regulations tighten, such as with the introduction of the Battery Passport in Europe by 2027, it will become imperative to not only identify but also understand the origins of defects throughout the battery lifecycle. The Battery Passport aims to provide detailed traceability and documentation of a battery’s materials and production history. This increased transparency requires manufacturers and stakeholders to adopt more rigorous quality control measures, ensuring that defects are not only noted but their causes are thoroughly investigated and documented. While defects may inherently remain unpredictable, the system is designed to ensure that once identified, their causes are traceable and well-defined. In short, failing to pinpoint the cause of defects will no longer be acceptable in light of the upcoming developments in the battery market. Such information can be utilised to optimise manufacturing processes thereby minimising the production of defective batteries.

Imagine a future where Li-ion battery production is as streamlined and defect-free as the precision-driven semiconductor industry. Just as the semiconductor metrology and inspection market, valued at $7.3 billion in 2021, revolutionised electronics manufacturing with meticulous quality control, implementing optimisation solutions in battery production could similarly transform the battery industry. By enhancing production processes, manufacturers can increase the throughput of battery production lines while significantly reducing costs and waste. This breakthrough would enable the scaling of Li-ion battery capacity and lower costs, driving high sales volumes and rapid market expansion, mirroring the rapid growth seen in the semiconductor sector.

Optimising battery manufacturing technology can enhance batteries for existing applications and unlock new uses and integration with other cutting-edge technologies, such as advanced solid-state battery management systems. This synergy could lead to more efficient power grids and smarter transportation systems, opening new avenues of possibility sooner than we might think.

Why are battery diagnostics so important?

Battery diagnostics have the potential to unlock unprecedented opportunities for implementing smart manufacturing practices and advanced process control solutions that can accelerate yield ramp, reduce scrap rates, and achieve higher production throughput. By gathering actionable insights on the cause, type, frequency, and impact of battery defects during production, manufacturers can inform mitigation processes to implement in the production line.

Optimising production can maximise investment returns on battery assets by minimising failure rates during operation, and therefore the need for frequent replacements, leading to cost savings. By increasing battery lifetime and concentrating failure occurrences in the battery value chain, diagnostics enable efficient planning for recycling, repurposing, and scrapping processes, thereby economically reducing environmental impact.

Ultimately, battery material variability often forces manufacturers to include additional cells in battery packs to mitigate the risk of poor performance. Based on discussions with manufacturers, many OEMs over-provision their batteries by adding up to 10% more cells than necessary, significantly increasing the cost and weight of the final product. By utilising advanced battery diagnostic tools to optimise production and reduce material variability, manufacturers can eliminate the need for these extra cells. We estimate this could result in savings of up to $40 billion by 2030, based on the projected market value of the battery manufacturing industry. The exact figure will depend on further reductions in cell prices and the realisation of gigafactories, which will largely depend on market demand in the coming years.

Battery Diagnostics 101:

Physical, chemical and electrical metrology sensors acquire manufacturing data during battery production. Subsequently, data-processing software enables manufacturers to make real-time decisions based on sensing data acquired.

Depending on the methodology, the extracted data can provide valuable insights such as intrinsic battery material properties and performance data such as voltage responses and charging status.

If battery diagnostics isn’t new, why are battery defects still a problem?

Battery manufacturers have integrated diagnostic tools into their factories for decades. These tools have been a cornerstone of R&D efforts aimed at optimising batteries and identifying novel designs since the inception of battery production. But as the heading suggests, we still have work to do, and here’s why…

Directly inspecting internal battery properties is challenging due to the typical encasement within stainless steel cans, plastic pouches, or hard prismatic shells. Existing methods that create visual images of internal batteries capable of inferring the nature and location of defects are drastically time-intensive and cost-prohibitive.

Given the thin profit margins In the battery production industry, there’s a preference for low-cost methods, often relying on indirect cycling measurements to assess numerous battery properties. However, these tools often come with reliability concerns as they use estimates rather than direct measurements. Should the data suggest the battery possesses defects, it can be nearly impossible to discern the nature and location of the defect in the cell and to take focused action to mitigate it.

I have a problem, but I don’t know what my problem is…

We need higher-resolution diagnostic tools that don’t break the bank

Indirect cycling measurements predominantly measure voltage, temperature, and current using hardware. Leveraging these parameters, software-based algorithms estimate the cell’s state of charge (SOC), resistance, and capacity. Subsequently, these estimations are used to deduce the battery’s energy and power metrics.

As analysis progresses from direct measurements to first and then second-order estimations, there is a noticeable decline in resolution. This directly affects the reliability of the data obtained, increasing the likelihood of mistakes in interpreting battery conditions and longevity. Consequently, defective batteries may infiltrate the market, leading to costly recalls.

Adding to these challenges, operators also tend to rely on using pre-existing lookup tables to characterise the diagnostics data acquired from battery testing, which can be inadequate in addressing the intricacies of different battery systems leading to misinterpreted battery analysis.

As multi-chemistry energy storage systems and novel battery archetypes such as BYD’s “Blade battery” become more prevalent, there is a pressing need for chemistry-agnostic, format-agnostic cross-web diagnostics tools that offer high-resolution.

We’ve talked about cost and resolution (accuracy and precision), another challenge is discerning where to integrate diagnostic tools in the production line to maximise profitability (I.e. “Get the best bang for your buck”).

Traditionally, battery producers have incorporated diagnostic tools at the end-of-line stages of production, providing a final check to ensure cell quality before market release.

However, critical micro and nano-scale material defects could accumulate from chemical processes that shape battery characteristics in the early stages of battery production. These defects are inherently probabilistic and can persist throughout later production processes, remaining undetected until the final quality control stage. Consequently, valuable materials and manufacturing time may result in defective cells destined for scrapping, a situation that could have been avoided if diagnostics tools identified defects early in the production process.

Detecting defects early in the production process is crucial, as it enables the timely removal of defective units and facilitates immediate corrective actions to maintain high-quality battery production. This proactive approach can significantly enhance yield and performance, thereby accelerating the achievement of target production rates. Traditional microscopy tools used to detect micro and nano-scale defects at early production stages (before battery assembly) — such as XRD, SEM, and X-ray CT — are not without drawbacks; they are inherently energy-intensive, require off-line inspection compromising manufacturing throughput, and cost-prohibitive, which hinder their widespread adoption in manufacturing environments.

Placing the diagnostic tool earlier in the production stages carries risks too, as defects may arise in subsequent stages and go undetected, potentially entering the market without detection. Therefore, regardless of early-stage diagnostics, we are confident that implementing a comprehensive diagnostic tool for final QA/QC at the end-of-line is essential before market rollout, ensuring that any latent defects are detected and addressed before products reach consumers.

Determining a single optimal location for diagnostic tools is less effective than a more dynamic approach, which recognises that the optimal inspection solution involves deploying multiple tools at different stages of the production line. Each tool is specifically optimised to measure a particular property of the battery, such as electrode consistency or slurry viscosity. This tailored approach allows for a comprehensive assessment of the battery’s attributes throughout production. Implementing versatile diagnostics that can be easily integrated into various segments of the production line without extensive adjustments — offering both mobility and plug-and-play capabilities — would provide producers with deeper insights and greater flexibility in monitoring and enhancing production quality.

Modern gigafactories have evolved to incorporate in-line real-time sensing tools for measuring at each processing step. This can include lasers for electrode thickness measurements, beta-rays for discerning mass loading and imaging systems such as in-line vision cameras. These tools are crucial for continuously monitoring and assessing battery quality throughout the manufacturing process.

Nonetheless, existing in-line tools often employ spot sizes that are too large (~1 mm) to detect certain defect types, cover a limited portion of battery material surface, involve cumbersome process integration requirements, are limited to indirectly measuring one property, and/or cannot measure certain chemistries.

Battery diagnostics solutions must prioritise several key factors: high resolution, cost-effectiveness, speed, non-destructive testing capabilities, and compatibility with various stages of production.

In mapping the diagnostics landscape, compelling next-generation diagnostics tools have been identified with the potential to provide such capabilities.

Other notable merits include multifunctionality, inline and real-time capability, adaptability to different battery chemistries, and a common software platform to streamline data collection and analysis.

The true potential of battery diagnostics lies in merging these tools, leveraging one method to enhance the capabilities of another. For example, utilising high-resolution X-ray CT imaging data to train lower-cost acoustic transmission measurements allows for the inference of similar findings at a reduced cost and in less time.

Where diagnostics providers can find their volume in battery production

Gigafactories are colossal facilities engineered to manufacture vast quantities of batteries. Due to their immense output, these facilities wield significant market influence.

Battery diagnostics can accelerate yield ramp-up and optimise on-tool processes, enabling new factories to achieve profitability more quickly thereby establishing their essential role in the production ecosystem.

Over 240 Li-ion battery gigafactories are currently in the planning or construction stages globally. A notable 82% of gigafactories are concentrated in China with most batteries produced by battery manufacturing goliaths like BYD and CATL. This concentration might dissuade diagnostics providers from targeting Western battery producers, given the perceived limited customer base in Western markets. This may also explain why most battery diagnostic providers service EV/ESS operators.

However, the number of gigafactories in the USA and Europe is expected to rise. The USA, EU and Japanese governments have all announced multi-billion dollar initiatives and subsidies to support Li-ion battery manufacturing and distribution. According to the US Department of Energy, 13 battery gigafactories are set to come online in the USA by 2025. Major automakers and battery manufacturers such as Ford, General Motors, and Stellantis are involved in these projects, with several factories established as joint ventures.

Across the pond, Europe is anticipated to host 27 gigafactories from 18 battery cell producers. Tesla and LG Chem are among the leading companies driving this growth, with gigafactories planned strategically across Germany, Sweden, and Poland​. Nonetheless, Northvolt, Europe’s battery champion, recently hit a bump in the road by halting expansion at their Swedish gigafactory and scaling back operations due to capacity issues and market slowdowns.

Despite the considerable gap in the number of gigafactories between Western markets and China, diagnostics providers have the potential to generate substantial revenue from individual clients due to the typical production volumes of gigafactories. Take Tesla’s Gigafactory in Nevada, a prominent example known for its immense scale. With a planned annual production capacity of 35 GWh in its initial phases, this facility could yield millions of batteries annually, according to battery size and energy capacity.

Failure to service the growing demand would leave the domestic battery industry dependent on imports and thus vulnerable to trade conflicts and supply shortages like those already seen with semiconductor chips; not to mention the huge carbon footprint of shipping battery cells around the globe that can pose a heightened risk to the industry due to increasingly stringent regulations around decarbonisation (battery passporting).

Battery diagnostics companies typically operate with a business model that includes an upfront acquisition fee for the purchase and installation of diagnostic equipment. This is complemented by recurring service contracts for software and AI-driven data analytics tools. These data analytics tools provide insights into process improvements and cause analysis, fostering strong customer adoption and ensuring steady recurring revenues. By offering quality assurance and quality control (QA/QC) services priced on a $/kWh basis, diagnostics providers could receive millions of dollars in revenue from such partnerships.

However, it should be noted that many planned gigafactories have faced delays, been paused, or even cancelled. Supply chain disruptions, particularly localised production challenges with lithium extraction and processing, have led to delays in several Li-ion battery projects across Europe. Moreover, the demand and sales of EVs have not grown as anticipated. This has complicated investment plans and led to the reevaluation of several gigafactory projects, with some experts predicting continued delays and uncertainty in future developments.

Maximising Venture Potential in Battery Diagnostics

Battery diagnostics servicing battery production represents an exciting investment opportunity. Over the last decade, $42 billion in venture capital and growth equity was invested in Li-ion battery companies in the USA. As battery manufacturing gains momentum, the diagnostics industry is experiencing complementary growth.

Startups are at the forefront of revolutionising battery manufacturing with novel diagnostic techniques that offer unparalleled insights into battery properties. Notable technologies that startups are developing include first-of-a-kind acoustic transmission methods, that enable real-time State of Charge (SOC) measurement, monitor solid electrolyte interphase (SEI) layer formation, and create detailed topographical maps of cell layers with exceptional precision.

The data generated by these advanced diagnostics provides opportunities for resource-efficient process optimisation at multiple stages of development and scale-up. This wealth of information serves as a foundation for implementing smart manufacturing practices such as digital twins and intelligent decision-making, driving the digitisation of the next generation of battery gigafactories.

Through these innovations, startups are playing a crucial role in reducing the economic and environmental costs of battery production, optimising the use of raw materials and energy, and minimising scrap rates of off-spec battery components.

Whilst exploring venture opportunities in this space, it is crucial to consider multiple factors:

1. Value Chain Integration: Assess each tool’s effectiveness in addressing specific customer pain points at different value chain segments and pinpoint the entry point with optimal gains for stakeholders.
2. High-Resolution Capabilities: Prioritise diagnostics technologies with higher resolution for richer performance insights. The greater the level of detail the tool can provide, the greater the potential for discovering new use cases compared to competitors.
3. Market Dynamics and Policy Trends: Navigate industry fragmentation and policy trends to discern evolving customer requirements. Stay updated on emerging policies such as battery passporting and their impact on battery manufacturing and user requirements.
4. Chemistry Adaptability: Evaluate diagnostics technologies that are adaptable to different battery chemistries. Investing in versatile technologies can enable market stability and sustained value across different battery types.
5. Beyond Batteries: Consider the broader applicability of diagnostics technologies beyond batteries. For instance, explore their potential in electrolysers and fuel cells for diversification and larger market opportunities.

Notably, there are various startups that are developing technologies that ensure compatibility for both inline and end-of-line QA/QC of assembled batteries before market rollout, preventing battery manufacturers from relying on multiple diagnostic tools. Tools lacking this dual compatibility can discourage manufacturers from subscribing due to the added complexity of integrating different diagnostics systems from independent providers, thereby increasing operational complexity.

Choosing a provider that offers both inline and offline products can offer several advantages:

1. Negotiate a fair price for a comprehensive diagnostic suite, covering both inline and offline offerings.
2. Combine findings from inline and offline tools to gain insights into optimising batteries based on properties detected at earlier manufacturing stages that impact later stages. This enables a better understanding of which defects and characteristics identified initially persist or evolve in the final product.
3. Simplify the learning process for operators by using a single diagnostic tool, reducing the additional cost and effort required for blue-collar workers to upskill on different software and hardware tools.

The search is on for that paradigm shift that will transform battery production and set a precedence for solutions to follow: start-ups that deliver diagnostic tools that are effectively a Swiss army knife, deployable for valuable insights at every part of battery production, seamlessly integrating into existing workflows and maximising value for stakeholders across the industry. If a solution can do this, while adhering to the key performance requirements mentioned earlier in this blog, it has the potential to become a jack-of-all-trades and a master of all.