SmaRackT - Smart Rack Monitoring

Object classification and quantification on the basis of near-field induction and machine learning

Initial situation

Material flows in production and logistics are often intransparent. Tracking and tracing requires manual, potentially faulty activities (e.g. scanning) or the cost-intensive tagging of articles and containers. These processes are very time-consuming and account for a significant share of total process costs, e.g. 25 to 35 % in intralogistics and spare parts supply. The automated tracking of material flows thus holds great potential in terms of process reliability and efficiency.

 

Objective

The research project SmaRackT (Smart Rack Monitoring) aims at the development and evaluation of an autonomous and efficient intelligent solution for the classification and quantification of objects at action points. The main characteristic is that none of the objects will be tagged, but identification and tracking will be enabled by inductive near-field detection and machine learning. Potentially, all material flows at action points along a process chain could thus be tracked. This could replace previous solutions such as barcode scanners.

 

Approach

To realize the research objective, the approach is split into three phases, which in turn are divided into a total of eight work packages (WPs).

Step 1 comprises a detailed analysis of application scenarios (WP 1), which, after final selection, leads to the deduction of requirements for the developed technology - e.g. regarding accuracy, range, size/weight of individual system components (WP 2).

Based on this, the technological development will be carried out in phase 2. First, a minimum viable product (MVP) is developed (WP 3). This basic solution is then technologically extended and further developed into a demonstrator (WP 4), whereby hardware components are adapted and the integration and application components of the software are refined. This particularly includes procedures for the recognition and comparison of objects (WP 5) and the integration into an own middleware platform (WP 6).

Practical application and evaluation of the demonstrator (WP 7) are the focus of phase 3.  Besides performance testing in the laboratory, also an installation under real conditions in cooperation with members of the project committee is carried out. Finally, the project is documented and the results are transferred.

Research partner

Project partners

  • Albrecht Jung GmbH & Co. KG
  • ambos.io GmbH
  • AUDI AG
  • Bernd Kraft GmbH
  • CIM Logistik Systeme GmbH
  • Computer OEM Trading GmbH
  • Cordes & Gräfe
  • Création Gross Gmbh & Co. KG
  • DigitEV GmbH
  • dmk Logistik Beratungs- und Beteiligungs GmbH
  • ELABO GmbH
  • FATH GmbH
  • FIS GmbH
  • KBS Industrieelektronik GmbH
  • Kühne + Nagel AG & Co. KG
  • Liv Tec GmbH
  • Miba Frictec GmbH
  • NeoLog GmbH
  • PickWerk GmbH
  • Pick to Light Systems S.L.
  • SSI Schäfer Automation GmbH
  • trilogIQa

Funding

The IGF project 21159 N of the Forschungsvereinigung Bundesvereinigung Logistik (BVL) e.V., Schlachte 31, 28195 Bremen, is funded by the AiF within the programme for the promotion of joint industrial research (IGF) of the Federal Ministry of Economics and Energy on the basis of a resolution of the German Bundestag.