Implement Localization Accuracy and Real-Time Metrics

Aberta
#759 0 comentários 0 reações 0 responsáveis Ver no GitHub

Mantenedores costumam responder em até 2 dias

Ninguém assumiu esta issue ainda.

Avaliação

Dificuldade
5/5
Tempo estimado
Mais de uma semana
Facilidade para iniciantes
15/100
Tipo de issue
Funcionalidade
Clareza
Razoavelmente clara
Status de atividade
Ativa
Stack de tecnologia
rust

Direção de pesquisa

The issue names no files, tests or entry points, so start by finding the existing localization pipeline and its task scheduling code in the repository. Decide the measurement boundaries and clock mappings before adding any instrumentation, since the issue requires each metric to have a documented definition, unit and clock domain. Done means synthetic traces with known delays and deadline misses produce the expected metric values, and the summaries can be reproduced from machine-readable output. This is design-heavy work, so discuss the approach with maintainers before writing code.

Escrita pelo modelo de indexação a partir do texto da issue.

Descrição

Implement consistent instrumentation and reporting for localization accuracy, scheduling delay, deadline misses, end-to-end latency, and input data age in the integrated Awkernel localization pipeline. Provide reusable metrics for concurrent replay evaluation and subsequent robustness experiments.

Scope:

  • Define measurement boundaries and clock mappings for all metrics
  • Measure position and orientation errors against timestamp-aligned ground truth where available
  • Measure scheduling delay from task readiness to execution start, distinguishing it from input-release delay and dependency waiting
  • Detect deadline misses using explicitly defined release times, deadlines, and completion events
  • Measure per-scan latency from pipeline input availability to the associated NDT result
  • Track when NDT and gyro odometry measurements are incorporated into EKF outputs
  • Measure the age of each consumed input from its measurement timestamp to consumption time
  • Export per-event records and aggregated summaries

Acceptance criteria:

  • Each metric has a documented definition, unit, clock domain, and applicable task or output.
  • Trace identifiers associate scans, task executions, measurements, and outputs without relying solely on timestamp proximity.
  • EKF output latency and freshness are reported separately for relevant input streams rather than attributed to an arbitrary latest scan.
  • Deadline reports include miss counts, evaluated-job counts, miss ratios, and tardiness; dropped or unfinished jobs are reported separately.
  • Timing summaries include median, tail percentiles, and maximum observed values with sample counts.
  • Accuracy reports distinguish ground-truth errors from differences against a reference implementation.
  • Synthetic traces with known delays and deadline violations validate the metric calculations.
  • Instrumentation overhead is measured and reported.
  • Machine-readable results and analysis commands are sufficient to reproduce the summaries.

Out of scope:

Replay workload generation, changes to localization algorithms or scheduling policies, and formal worst-case timing guarantees.

Linguagem predominante
Rust
Estrelas
282
Forks
12
Merge médio
2d 9h
PRs com merge (30d)
28

Preparar o ambiente

Este projeto não oferece contêiner de desenvolvimento, Dockerfile nem guia de contribuição, então a configuração fica por sua conta: comece pelo README e veja nosso guia da primeira contribuição para os passos gerais.

Primeiros passos

  1. Leia a issue inteira e depois o guia de contribuição do projeto.
  2. Comente na issue dizendo que vai assumir — evita que duas pessoas façam o mesmo trabalho.
  3. Faça um fork do repositório e trabalhe em uma branch.
  4. Abra um pull request que referencie o número da issue.

Mais de tier4/awkernel

Todas as issues de tier4/awkernel

Issues semelhantes

Mais issues de Rust

Receba novas issues na sua caixa de entrada

Um resumo curto de issues do GitHub para quem está começando.