Student project: Real-Time, Robust, Adaptive Indoor Localization through BLE and Lidar Sensor Fusion

Indoor localization is integral to applications ranging from autonomous mobile robotics and warehouse automation to smart building management and personalized location-based services. However, achieving real-time, robust, and accurate localization indoors is challenging due to complex layouts, dynamic changes (moving furniture, varying crowd density), and signal degradation. Current methods often rely on a single sensor type, e.g. IMU, Bluetooth, UWB, Radar, Lidar. In this thesis, the focus will be on the combination of RF solutions (Bluetooth Channel Sounding) and light-based solutions (Lidar).

What you will do

• Literature Review and State-of-the-Art.
• Bluetooth Channel Sensing:
  Studies show that Angle-of-Arrival (AoA) estimation and Channel State Information (CSI) can significantly improve indoor positioning beyond RSSI-based methods. Existing works demonstrate decimeter-level accuracy in controlled conditions, though often with increased complexity. Some approaches apply machine learning to predict location from channel features efficiently.
• IMU sensing.
• LiDAR-Based SLAM and Feature Extraction:
  LiDAR-based methods achieve high spatial precision using SLAM algorithms like Hector SLAM or Cartographer. Advanced feature extraction (line segments, corners, occupancy grids) and semantic labeling approaches improve robustness. However, these methods can be computationally expensive. Efficient feature selection and incremental update strategies can enable near real-time performance.
• Latent Representation Learning (LatentSLAM):
  LatentSLAM methodologies show that learned compact embeddings can represent places more robustly than raw sensor data. Such latent spaces facilitate quick place recognition, map updates, and adaptability to changing conditions. Tailoring these techniques for multi-modal (RF + LiDAR) data fusion can provide a powerful, compact state representation that supports real-time inference.

Research Objectives:

Latent Space Fusion Framework:
Design and implement a computationally efficient framework that merges BLE RF signals and LiDAR features into a shared latent space. Ensure that the embedding process, either via a lightweight neural network architecture or a statistically robust dimensionality-reduction approach, runs at or near the sensor rate.**

Efficient Inference and Filtering:
Employ Bayesian filters or pose-graph optimization using latent space measurements. Alternatively, lightweight neural networks can directly predict positions from latent representations. Each approach will be optimized for low-latency processing on embedded hardware, targeting execution times compatible with navigation requirements (e.g., 10 Hz or better).

Performance Evaluation under Dynamic Conditions:
Test the system in various indoor scenarios: offices with moving furniture, corridors with passing people, Measure localization accuracy, update rate, latency, and robustness over time. Compare against baseline RF-only, LiDAR-only, and static fusion methods to quantify improvements in performance and adaptability.

What we do for you

You will be working on cutting-edge research on a topic that is relevant to both academic and industrial research groups. To help you in this journey, we offer a flexible environment where you can be the leader of your own research while at the same time having the support of experts to complete your tasks. As part of the team in IMEC-Netherlands, you will have opportunities to learn from some of the best minds. 

Who you are

• You are a MSc student in Electrical Engineering or Embedded Systems
• You are available for a period of at least 9 months.
• Prior experience and affinity with wireless signal processing.
• You are familiar with MATLAB, Python, AI Tools.
• Experience in C is a plus
• You are entitled to do an internship in the Netherlands.
• You are self-starter and able to work independently.
• Good written and verbal English skills. 

Interested

Does this position sound like an interesting next step in your career at imec? Don’t hesitate to submit your application by clicking on ‘APPLY NOW’.
Should you have more questions about the job, you can contact jobs@imec.nl.