Energy

Battery Host System Controller

Dynamic Grid - Energy Storage

Download Case Study
A desk with a computer, network device, cables, technical diagram on energy storage, notebook, and coffee mug in an office setting.
Executive Summary
Key Challenges
Solution
Outcomes
End Results

At A Glance:

120ms real-time protocol translation

continuous Modbus–CAN bus bridging between Microgrid Controller and BMU fleet

Vendor-agnostic interoperability

complete SunSpec compliance (Models 1, 802, 803, 804) removes Microgrid Controller lock-in

Zero physical hardware required for development

full emulator-driven workflow (BMU, MgC, Fieldbus, DC3 emulators) enabled parallel software and hardware tracks

About The Client:

Dynamic Grid is a Maine-based clean energy company developing commercial Energy Storage Solutions (ESS) designed to work alongside microgrid controllers to provide safe, intelligent management of large-scale battery arrays. Their ESS product integrates Toshiba Battery Management Units (BMUs) — each controlling up to 50 Cell Management Units — into containerized battery installations capable of supporting 2 to 20 BMU racks per deployment.

Executive Summary

Dynamic Grid needed a production-grade software bridge to connect its Energy Storage Solution to any SunSpec-compliant Microgrid Controller while safely orchestrating up to 22 Toshiba Battery Management Units in real time. VividCloud designed and built the Battery Host System Controller (BHSC), a Python-based embedded application that translates between Modbus/SunSpec and CAN bus protocols every 120 milliseconds, implements all nine Toshiba safety procedures, and enforces independent failsafe contactor control, all without any cloud dependency. Delivered over an 8-month engagement spanning assessment and full solution engineering, the result is a lean, vendor-agnostic control system, validated by a custom emulator suite and 40+ automated tests, that lets Dynamic Grid bring its containerized battery storage product to market with any compliant Microgrid Controller and full confidence in its safety architecture.

Key Challenges

Dynamic Grid faced a demanding set of technical and operational pressures:

  • Real-time protocol incompatibility between the site’s Microgrid Controller, which speaks SunSpec Modbus, and the Toshiba Battery Management Units, which communicate over a proprietary CAN bus protocol, with no existing bridge between the two.
  • Safety-critical complexity across nine distinct operational procedures (start, stop, power failure, overcharge, temperature faults, and more) that had to be implemented faithfully and independently of the Microgrid Controller in case of failure or silence.
  • No physical hardware available for continuous development and testing, requiring a way to validate the full system before BMU and controller hardware were consistently on hand.

The organization needed a partner with both embedded software depth and electrical engineering expertise to deliver a vendor-agnostic, production-ready control system without locking their product to a single Microgrid Controller.

Solution

Battery Host System Controller (BHSC) Software

The BHSC is a Python application running on an Ubuntu-based industrial SBC. It is the ESS ‘control brain,’ continuously monitoring both the MgC and the BMU fleet and acting as the authoritative translator between them.

Infographic showing workflow and features of a Battery Host System Controller for dynamic grid energy storage and management.

Key architectural components include:

  • Modbus Server (SunSpec): The BHSC presents itself to the MgC as a SunSpec-compliant Modbus device, populating and maintaining holding registers for Models 1 (Common), 802 (Battery Base), 803 (Li-ion Battery Bank), and 804 (Li-ion Strings). Multiple Modbus device instances are used to represent the hierarchical ESS structure.
  • CAN Bus Interface: A dedicated CAN bus layer manages bidirectional communication with up to 22 Toshiba BMUs across one or more CAN channels, parsing proprietary message frames (Frame IDs 0x00–0x0F and 0x10–0x18) and translating them to SunSpec register values.
  • BMU Shadow System: Inspired by AWS IoT device shadow concepts, each BMU is represented in-memory by a ‘shadow’ that holds the most recent copy of each CAN message type received. This decouples real-time CAN bus reception from the SunSpec calc cycle, enabling efficient, thread-safe access to current BMU state.
  • SunSpec Calc Cycle: A one-second calculation engine reads BMU shadows and applies business logic to populate SunSpec fields — averaging SOC across BMU strings, aggregating current and power, computing state of health, and deriving charge status, control mode, and event flags.
  • Procedure Engine: Nine distinct Toshiba procedures are implemented as discrete state-machine modules: Start, Stop, Power Failure, Restoration, High Temperature, Over Temperature, Overcharge/Overdischarge Alert and Fault, Current Leak Detection, Cell Balance, and Time Adjustment.
  • Failsafe Logic: Independent of MgC directives, the BHSC monitors for conditions that require immediate contactor intervention — including overcurrent, overvoltage, significant over-temperature (>54°C), and loss of MgC heartbeat.
  • Fieldbus Integration: A separate Modbus connection manages the switchable 12V BMU power distribution system via a Fieldbus subsystem.
  • DC Meter Integration: An additional Modbus connection to an external DC voltage meter enables independent reporting of DC bus voltage.

Outcomes

  • Delivery: Completed full assessment-to-production engineering cycle in approximately 8 months, moving from requirements discovery through a production-ready Battery Host System Controller.
  • System Reliability: Multi-layered safety architecture, combining Toshiba procedure compliance, independent failsafe contactor control, and Microgrid Controller heartbeat monitoring, protects the battery array against a broad range of failure conditions without relying on a single point of control.
  • Quality: Validated through 40+ unit test modules and a full system integration suite, exercised against a custom emulator stack (BMU, MgC, Fieldbus, DC3 meter) so defects could be caught and resolved before physical hardware integration.
  • Performance Gains (from the BHSC):
    • Real-time protocol translation, with BMU status polled every 120 milliseconds for continuous, low-latency visibility into battery state.
    • Vendor-agnostic interoperability, since full SunSpec compliance means any compliant Microgrid Controller can integrate without software changes.
    • Reduced hardware dependency risk during development, as the emulator-driven workflow let software and electrical engineering tracks proceed in parallel.

End Results

Unlike many embedded systems vendors that focus narrowly on protocol implementation, VividCloud applied the full rigor of modern software engineering practice to the Battery Host System Controller program. By grounding the architecture in operational safety and real-world deployment constraints, the team ensured every design decision, from the BMU shadow system to the procedure engine, served Dynamic Grid’s actual product requirements rather than a generic interpretation of the spec. The emulator-driven development workflow, paired with 40+ unit test modules and a full system integration suite, created a safety net that surfaced defects in software long before physical battery hardware was ever connected, substantially lowering the risk of late-stage integration surprises that often plague embedded hardware/software projects.

Equally important, VividCloud partnered with Dynamic Grid in a capacity that extended well beyond writing code. By contributing electrical engineering expertise across SBC selection, power distribution design, and cable harness layout, and by building a Modbus/CAN translation layer that keeps Dynamic Grid vendor agnostic on the Microgrid Controller side, VividCloud helped Dynamic Grid de-risk both the technical and commercial path to market. The result was not only a production-ready, safety-compliant control system, but also a foundation Dynamic Grid’s own team can maintain and extend without dependency on cloud services or a single integration partner. This approach reflects VividCloud’s role as more than a contract developer: we are a partner that turns complex, hardware-adjacent software challenges into a durable, ownable product.

Download Case Study

Related Client Success Stories

A warehouse with conveyor belts transporting packages. The area is equipped with various machinery and material handling equipment, including Dematic robotic arms for handling items. Yellow railings and safety barriers are present, creating an organized and automated space.
Micro-Fulfillment the Modern Way with VividCloud
With a growth rate of 21.2% in 2021, Dematic was listed as the world’s second-largest materials handling systems supplier, with a revenue of $3.2 billion. Dematic has research and development engineering centers, manufacturing facilities, and service centers located in more than 35 countries around the world.
Industrial
View Project
View All Projects
Two people monitor computer screens in a Dematic Control Tower, overseeing a busy warehouse floor filled with shelves and workers.
Dematic’s Control Tower
With a growth rate of 21.2% in 2021, Dematic was listed as the world’s second-largest materials handling systems supplier, with a revenue of $3.2 billion. Dematic has research and development engineering centers, manufacturing facilities, and service centers located in more than 35 countries around the world.
Industrial
View Project
View All Projects

Ready to have us on your team?

With VividCloud, you get ingenuity on demand to solve your most pressing cloud software engineering challenges. Drop us a line to begin the conversation — we can’t wait to hear from you.

Send us a message.

Subscribe
Privacy Policy