Automated Inspection of Shipping Containers at Container Depots

Background

There are about 11 shipping container depot operators in Singapore, operating around 20 container depot yards. At each of these yards, the inspection of containers (20 ft and 40 ft) for defects is carried out manually by a surveyor, in accordance to standards such as the Technical Reference 39 (TR 39 [1], developed by the Singapore Standards Council, Logistics Technical Committee), Cargo Worthy Standards by Shipping Lines or relevant IICL Standards (Institute of International Container Lessors). On average, about 3 to 5 surveyors are deployed to inspect containers at each of the container depot yards.

Existing Process

On average, about 100 to 200 containers are inspected daily for defects or damages. Generally, about 30 to 40% of the containers are expected to have some form of defects or damages that should be detected and identified.

The existing inspection process is carried out by visual survey on all visible areas of the container, interior and exterior. The estimated time taken to inspect each container may range from 5 to 10 mins (no defects), 20 minutes (minor defects) and up to around 40 minutes (serious defects). Inspection is typically carried out under daylight with containers mounted on trailers. Common defects or damages may include dents, deformations, or visible spillages of liquids.

Further details on the reporting criteria for conditions of containers can be found in TR 39 [1].

As the inspection may include some work at heights, the appropriate safety considerations for fall prevention would have to be put in place. Other challenges include inspection of the undersides of container surfaces which may be obscured from view unless the surveyor accesses under the trailer chassis.

The solution sought should be able to inspect the container autonomously or remotely, effectively detect, assess and classify the defects on containers and tag the findings to the containers based on their identification numbers. Ideally, the number of surveyors required per inspection yard should be reduced to 1 or 2.

Requirements

There are no restrictions to the types of technology deployed, including drones, gantries, laser scanners, computer vision, AI, AR etc. However, the proposed solution should address the following requirements:

  • Efficiently detect, assess and classify the defects, damages or uncleanliness on containers in accordance to the reporting criteria specified in TR 39, Cargo Worthy or IICL Standards
  • Offer significant time and manpower savings over the existing manual inspection process
  • Be able to inspect defects at heights, and at all exposed surfaces of the containers
  • Be able to provide location information of the defect position and recommend the required action to rectify the defect
  • Allow the data collected to be easily integrated with existing container depot management systems
  • Be able to identify and recognise container identification numbers printed on the sides of the containers
  • Be able to operate remotely, or autonomously
  • Be able to detect defects from a range of container surfaces and dimensions (20ft, 40ft, ISO tanks, reefers etc.)
  • Be easily deployable with minimal infrastructure costs
  • Achieve a minimum accuracy of 70%

Data and physical containers for the training of AI models will be made available to shortlisted proposals.

Geographical Restrictions

No geographical restrictions on the origin of the technology. However, the proposed solution should be deployable for test and demonstration on-site at a Singapore-based container yard.

Minimum Required Technology Readiness Level (TRL)

Level 7

 

Desired Outcome

The solution should result in significant time savings and reduce the reliance on manual labour for the container inspection process, hence improving the depot’s handling capacity. The solution should also reduce/eliminate risky activities such as working at height or movement under the container trailer chassis. Being able to remotely/autonomously inspect the containers will make the job more attractive too.

 

Development Timeframe

6 months (including prototype and trial).

 

 

Further Details

Further details of the challenge statement may be found at the Briefings Page, in the form of video recordings of the virtual briefing sessions, Q&A transcripts and presentation slides if available.

 

 

Instructions on Submitting Proposals

Download the proposal submission form in Microsoft Word format at the 'Attachment' section of the Challenge Statement Listing Page and fill in the details of your project proposal. You can then proceed to the proposal submission page by logging in and clicking the 'Submit Proposal' button on the right side of this page, where you will have to fill in the form fields and attach the filled Microsoft Word project submission form before submitting. Do also upload supporting documents/files that may support your proposal.

 

Challenge

TCC Industry

Organisation

Container Depot & Logistics Association of Singapore (CDAS)

Themes

Big Data, Data Analytics
Robotics and Automation
Process Enhancements

Proposal submissions are open from 22 Jul 2020 to 15 Oct 2020