Internship: Research and Validation of Battery SoC and SoH Estimation for Autonomous Robots

Job Description

Electrification plays an increasingly important role in Lely autonomous robot systems. Battery behaviour, lifetime, cost and diagnostic reliability directly influence robot availability and future product choices. Within this internship you will contribute tothe research and validation of practical methods to estimate the state of charge (SoC) and state of health (SoH) of Lead Carbon batteries used in robot applications.

The assignment focuses on comparing estimation approaches across different system cost and complexity levels. You will startfrom a very low-cost setup, where there is no battery communication and only limited or no direct battery measurementsavailable. From there, you will evaluate increasingly capable solutions, up to a battery system with a cell balancer or batterymanagement electronics that can communicate battery data to the robot.

The goal is to determine which SoC and SoH estimation approach is technically suitable, economically sensible and realistic to integrate in one of our robots. The assignment combines theory, data analysis and experimental validation.

Part 1 - Analysis of SoC and SoH estimation methods

You start with a structured comparison of battery estimation methods based on literature research, supplier information and technical documentation. The analysis should include both low-complexity and higher-complexity options. Topics you investigate include:

• State-of-charge estimation by voltage, current integration / coulomb counting and model-based approaches
• State-of-health indicators such as available capacity, internal resistance, voltage behaviour and ageing trends
• Impact of limited sensing: no communication, limited measurements or indirect robot-side measurements
• Value of additional electronics such as a cell balancer, BMS or communication interface
• Cost, complexity, robustness and integration effort for each method
• Expected suitability for lead-carbon batteries and other relevant robot battery options

The outcome of this part is a technically supported overview of the possible estimation methods, including their strengths,limitations and realistic usability in a robot environment.

Part 2 - Experimental validation and comparison

Next to the theoretical comparison, you validate selected methods experimentally. You will set up and execute battery teststhat allow the different SoC and SoH estimation approaches to be compared using real measurement data. The experimental work may include:

• Charge and discharge cycles to characterise battery behaviour over time
• Testing under different temperature conditions, using a temperature cell or climate chamber where available
• Comparison of new batteries with aged batteries from the field
• Data logging and analysis of voltage, current, capacity, efficiency and temperature-related behavior
• Validation of estimation accuracy versus measured reference data

You will analyze the recorded data and compare the practical results with the theoretical expectations. This should lead to a clear recommendation on which estimation approach gives the best balance between cost, complexity and diagnostic value fora robot application.

Part 3 - Recommendation for robot application

Based on the research and experiments, you translate the findings to a practical recommendation for use in a Lely robot. Thererecommendation should explain what can be achieved with a very low-cost system, where extra measurements orcommunication become valuable and which solution is most likely to fit current or future robot platforms. Depending on the planning and available hardware, the assignment may also include a small proof of concept in which thechosen method is implemented or demonstrated using Python-based data processing, a test setup or robot-relevantmeasurement data.

Practical information

• This is an internship assignment with a strong research and validation component.
• Start date in consultation.
• Preferred duration: 24 weeks.
• You are available at least 32 hours per week.
• You will mainly work from the Lely Campus in Maassluis.
• You will receive guidance from engineers in Electronics, Embedded Software and System Engineering.

We are looking for an…