Robot Batteries: A New Growth Play for SG Startups

EV demand is wobbling, but battery innovation isn’t slowing down—it’s looking for the next “always-on” market. South Korea’s biggest battery makers are already making that bet: robots.

Nikkei Asia reported this month that LG Energy Solution and Samsung SDI are actively pushing into batteries for service robots, autonomous mobile robots (AMRs), and humanoids as the EV market cools and price pressure rises. The interesting part for Singapore founders isn’t the Korea-versus-Korea competition. It’s the signal: robotics is becoming a battery-led category, and that changes how products get built, funded, and brought to market across APAC.

This piece sits within our “AI Business Tools Singapore” series, because robotics is no longer “just hardware.” It’s an AI deployment channel. And when you ship AI into the physical world, battery performance becomes a product feature—not a spec sheet detail.

Why South Korea is pivoting from EV batteries to robot batteries

Answer first: EV batteries are a scale business under margin pressure; robot batteries are a performance business with higher prices and deeper customization.

Korea’s battery giants are facing a tough reality: EV growth can pause, incentives can change, and competition can crush pricing. Nikkei cited LG Energy Solution’s operating loss of 122 billion won in Q4 even after 332.8 billion won in U.S. production subsidies, and Samsung SDI’s 299.2 billion won operating loss in the same period.

Robotics offers a different structure:

• Lower volume than EVs, but higher willingness to pay for performance.
• More customization, because each robot form factor creates unique constraints.
• Stickier customer relationships, since battery packs and control software get co-designed.

Nomura’s estimate (as cited in the article) is the cleanest takeaway for strategy teams:

Robot batteries could sell for roughly $200–$350/kWh by 2030, versus $80–$120/kWh for EV batteries.

If you’re a startup, that’s a familiar pattern: the early market isn’t huge, but margins and differentiation are.

What makes robot batteries different (and why it matters to AI-driven robotics)

Answer first: Robots punish batteries through frequent cycles, fast bursts of power, and strict weight limits—exactly where high energy density and durability become differentiators.

Robots are “battery-constrained AI products”

A modern robot isn’t a motor with a battery. It’s a mobile compute stack: sensors, perception, edge inference, connectivity, and actuation. Humanoid robots amplify this, because you’re powering bipedal locomotion plus onboard AI.

That’s why the battery requirements look more like aviation and industrial tools than passenger EVs:

• High power density (bursty loads for movement)
• High energy density (runtime without adding weight)
• Fast charging (shorter downtime)
• High cycle life (frequent duty cycles)
• Compact packaging flexibility (odd shapes, tight spaces)

Nikkei notes LG is supplying cylindrical battery cells to multiple global robot makers—cells that already show up in power tools and electronics—because they fit the “small but powerful” requirement.

The overlooked point: battery choice shapes the whole robot

Founders often treat batteries as a procurement line item. In robotics, battery selection shapes:

• Mechanical design (weight distribution, stability)
• Thermal design (cooling, safety)
• AI performance (compute power budget)
• Commercial model (swappable packs vs charging docks)

If your robot is part of an AI business tool (e.g., warehouse picking, cleaning, delivery, inspection), uptime is your revenue. Battery decisions directly influence unit economics.

The market size is smaller than EVs—so why should Singapore care?

Answer first: Because Southeast Asia’s first wave of robotics adoption will be practical, service-led, and battery-limited—exactly where Singapore companies can build defensible niches.

Nomura (per the article) expects robot batteries to reach 1–3 GWh by 2030, compared to 1,647 GWh for EVs and 750 GWh for energy storage systems. So no, this isn’t a “replace EV” story.

It’s a new profit pool story.

Here’s where Singapore fits in.

Southeast Asia adoption will be “unsexy” and high-frequency

Humanoids grab headlines, but the region’s near-term demand tends to come from work that’s repetitive and operationally painful:

• Facilities management: cleaning, security patrols, inspection
• Logistics and warehouses: AMRs, pallet movement, inventory scans
• Healthcare operations: delivery robots, support tasks in hospitals
• F&B and hospitality: service robots, kitchen runners, tray return

Singapore is already a testbed for these deployments because customers are willing to pay for reliability and compliance. That makes it a strong launchpad for founders planning regional expansion to Malaysia, Thailand, Indonesia, and Vietnam.

Battery performance becomes a go-to-market advantage

Most robotics go-to-market plans over-focus on demos and under-focus on operations. The pitch that wins isn’t “our robot uses AI.” It’s:

• “We guarantee X hours runtime.”
• “We can recharge in Y minutes.”
• “We hit Z tasks per shift without battery swap.”

When your commercial contract includes uptime and SLAs, you’ll care a lot about cell quality, pack engineering, and charging strategy.

Partnership playbook: how SG startups can work with Korean battery leaders

Answer first: Treat Korean battery makers as strategic partners for performance, certification, and scale—not just suppliers—and pair that with local AI, integration, and distribution.

Nikkei highlights two models:

• LG Energy Solution: already supplying cylindrical cells to multiple global robotics firms, discussing next-gen specs and mass production timelines.
• Samsung SDI: collaborating via an MoU with Hyundai Motor Group to co-develop robot-optimized batteries (Hyundai owns Boston Dynamics).

For Singapore startups, the practical approach is to pick a collaboration depth that matches your stage.

Stage 1 (MVP): design for availability, not perfection

Early on, supply chain risk kills robotics pilots. Use a battery architecture that can evolve:

• Standardized cell formats where possible (often cylindrical)
• Modular pack design for serviceability
• Conservative thermal margins (Singapore’s ambient heat is unforgiving)

Your goal is to prove unit economics and reliability before you chase the “perfect pack.”

Stage 2 (Pilot-to-scale): co-design the pack, own the battery data

This is where strategic partnerships matter. A credible battery partner helps you:

• Improve cycle life and fast-charge performance
• Navigate safety certifications and compliance
• Build a roadmap for mass production

Just as important: own your battery telemetry. Battery health is a predictive maintenance input.

If you’re building an AI operations layer, battery data becomes part of your moat:

• Remaining useful life prediction
• Charge optimization by route/shift
• Anomaly detection for safety events

Stage 3 (Regional expansion): productize charging and service

In Southeast Asia, the robot that wins is the one with the simplest operations:

• Charging docks that tolerate messy environments
• Swappable battery programs with clear SOPs
• Field-service friendly design

Battery makers can support scale, but you must design for service reality.

What to build next: robotics + AI business tools that sell in Singapore

Answer first: The strongest opportunities combine a clear operational pain point, a repeatable deployment model, and battery-aware AI that improves uptime.

If you’re looking for actionable directions, here are four that I’d bet on for 2026 planning cycles.

1) Battery-aware fleet ops software (the “battery brain”)

Most fleets still treat battery as a percentage indicator. Build software that predicts and schedules:

• Charging windows
• Task allocation based on battery health
• Degradation-aware routing

This is classic AI business tools Singapore territory: operational AI that directly reduces cost per task.

2) Fast-charge infrastructure for indoor robotics

Robots often charge where people work: malls, hospitals, warehouses. That creates constraints:

• Safety and thermal management
• Space limitations
• Electrical load planning

There’s room for startups that package charging as a system: hardware + monitoring + maintenance.

3) Vertical robots with guaranteed SLAs

Pick a vertical where uptime matters and ROI is measurable (warehouses are the obvious one). Win with a commercial offer tied to outcomes:

• “Cost per pick”
• “Cost per cleaned square meter”
• “On-time internal deliveries”

Battery reliability is the hidden enabler of those SLAs.

4) Retrofitting and second-life programs (early, but real)

As robot fleets grow, battery replacement becomes a lifecycle business. Startups can build:

• Battery diagnostics and grading
• Refurbishment workflows
• Second-life use in stationary micro-storage

This won’t scale overnight, but the economics get attractive once fleets hit meaningful numbers.

The hard truth: robotics won’t save every battery business—so plan accordingly

Answer first: Robot batteries are a premium niche for the next few years, not an immediate earnings rescue, which means startups should avoid “market size” narratives and focus on ROI narratives.

Analysts quoted by Nikkei warn robotics is unlikely to materially lift sector earnings within three years. That’s believable. Robotics adoption is operationally complex, and procurement cycles can be slow.

For founders and growth teams in Singapore, the right stance is:

• Don’t pitch robotics as “the next EV.”
• Pitch robotics as “the next high-margin, high-reliability deployment channel for AI.”

Robots will be bought by operators who care about uptime, safety, and service response time—not by people chasing hype.

Where this leaves Singapore startups in 2026

South Korea’s pivot is a reminder that when a sector matures, leaders look for adjacent categories where their strengths still matter. Battery makers are doing it. Singapore startups should, too.

If you’re building AI business tools in Singapore, robotics is one of the cleanest ways to turn AI into measurable operational outcomes. Just don’t treat the battery as an afterthought. In real deployments, battery performance is product performance.

What would happen to your sales cycle if you could confidently promise runtime, charging, and battery lifecycle costs—before the pilot even starts?