Smart Ports: Computer Vision

Part
01
of five
Part
01

Smart Ports: Computer Vision Insights

Computer vision helps smart ports mitigate inefficiencies, enhance security and safety, and assist with maritime side operations by facilitating applications for container identification, automatic object recognition, collision prevention, container placement/loading, and port security/surveillance. A deep dive into these findings has been provided below.

Computer Vision Application for Container Identification

  • Serial numbers serve to provide basic information about shipping containers arriving at ports, which is then used to help manage terminals, yards, and customs. Computer vision can be used to automatically identify these numbers.
  • By collecting this information via computer vision, logistics can be automated in real-time.
  • The computer vision technology can analyze things such as container numbers, the integrity of the container's seal, the state of the container and door lock, as well as the packing position of the container so that it can be placed on truck beds automatically while still maintaining the correct orientation of the container's door.
  • This application "can facilitate automated management, reduce manpower, and improve the container operation efficiency." It relies on the use of a camera and laser sensor.

Computer Vision for Automatic Image/Object Recognition at Ports

  • Images captured by cameras can be analyzed using computer vision in order to identify objects.
  • This computer vision application can be used in shipping ports to increase the number of inspections and optimize maintenance, especially preventative maintenance.
  • It can also be used to analyze the number and types of vehicles coming and going from the port and monitor the flow of traffic.

Computer Vision for Ship Collision Prevention

  • Computer vision technologies can be used to help improve the anti-collision capabilities of ships.
  • This technology uses data to reconstruct a visual image of the ship's environment and can asses the likelihood of a collision.
  • It can determine where other ships are and what types of ships they are, as well as assess bridge clearance and the amount of water between the ship and the shore.

Computer Vision for Container Placement on Ships

  • Computer vision can work virtual markers and unique container IDs to help "detect the position of a particular container on the ship" which is beneficial in a number of cases, including emergency situations.
  • This technology works using connected systems that communicate and share data with each other every time a container is loaded onto a ship.
  • The system can also work alongside machine learning algorithms to help determine the order in which containers should be loaded based on the destination of the shipment as well as the container's contents, and can also keep track of how much weight allowance is left on ship and how much available space there is (information that can then be sent to the crane master to help inform decisions).

Port Security and Surveillance

  • Computer vision technologies can work alongside GPU technology to turn basic security camera systems into intelligence surveillance systems for shipping ports.
  • Shipping ports are commonly at risk for asset destruction and reputational risks that can be hindered by security threats such as theft, smuggling, data breaches, and terrorist attacks. These threats can lead to disruptions in the flow of trade, large economical losses and even casualties and environmental damage.
  • Computer vision technologies can serve as surveillance systems which can help identify security issues in real-time while facilitating rapid intervention. These systems are more advanced than traditional video surveillance cameras (which usually only serve to secure against basic intrusions) and instead are able to count, track, detect anomalies, and even analyze behaviors.
Part
02
of five
Part
02

Smart Ports Implementation Costs

Implementing a smart port can cost upwards of $1-2 billion. The computer vision portion of those costs has the potential to represent ~29% of spending. More details about these costs are presented below.

Costs

  • Smart ports, or automated ports, require "massive capital expenditures." According to a news article published by Transport Topics, constructing an automated port "typically costs well over $1 billion." Peter Schneider, the executive VP at T.G.S. Transportation says that the total investment "can be in the $2 billion range."
  • According to McKinsey, setting up a smart port involves very high costs upfront. They estimate that in order to justify the investments, the operating expenses would need to be ~25% lower than those of a conventional port. Alternatively, the astronomical costs could be mitigated by rises in productivity (ideally by 30%, especially if operating expenses fell by 10%).
  • TechTalks reports that spending in the retail industry on AI technologies like computer vision could climb to $34 billion. Because computer vision facilitates inventory reporting, contributes to analytics, and empowers automation, 29% of spending on AI technology is expected to go to computer vision.
  • Computer vision used in automated image tagging is estimated to cost $0.01 per image.

Additional Related Information

  • Companies specializing in computer vision and smart ports include ABEJA (Japan; uses video analytics combined with deep learning to improve port operations), Delvify (Hong Kong; helps maritime enterprises categorize and analyze their data to optimize operational processes), and Langzou (China; uses networked cameras coupled with artificial intelligence capabilities to detect surrounding vessels or other obstacles).
  • "A smart port may be defined as a fully automated port where all devices are connected via the so-called IoT Smart Port. A network of smart sensors and actuators, wireless devices, and data centers make up the key infrastructure of the smart port.... Various sensors such as inertial sensors, ultrasonic sensors, eddy current sensors, radar, LiDAR, imaging sensors, and RFID readers and tags are used to collect the required data in order to transform the 'port' into a 'smart port.'"
  • Technologies that are heavily related to enabling smart and digitalized applications in the freight transport and logistics sector (such as digital rail, smart motorways, and smart ports) include cloud computing, IoT, social media networks, AI, big data analytics, immersive technologies, and distributed ledger technology (DLT).
  • The global smart ports market is expected to reach approximately $2 billion by the end of 2025. Major smart ports include the "Port of Rotterdam, Port of Hamburg, Port of Singapore, Port of Shanghai, Port of Los Angeles, Port of San Diego, and Port of Shenzhen." The market is also segmented into seaports (passengers and cargo shipment) and cargo ports (bulk ports).
  • Major vendors in the smart port market include Trelleborg, IBM, ABB, Traxens, Siemens, Liebherr, AGT Group, Kerry Logistics, Huawei, Cisco, Nokia, Ericsson, Konecranes, Navis, Kalmar, ProDevelop, Inform, and Wärtsilä. The global smart ports market will grow at a CAGR of ~18% between 2019 and 2025.
  • A different industry report noted that the smart ports market (by technology) was expected to reach $5.3 billion by 2024 and that it was currently valued at $1.7 billion (CAGR of 25%). The Asia Pacific market (China, Singapore, Japan, and Hong Long) is expected to be the largest market between now and 2024.
Part
03
of five
Part
03

Computer Vision Case Studies

Use cases for computer vision for automating port operations and enabling smart ports include automated anti-sway ship-to-shore container cranes, situational awareness technology for container vessels, and shipping container recognition solutions. A deep dive into these findings has been provided below.

Automated Anti-Sway Ship-To-Shore Container Cranes

  • This system was developed by MicroView & LingZhi Image Tech Corporation.
  • These companies partnered together to develop an "automated ship-to-shore anti-sway system that incorporates machine vision."
  • The system helps to regulate the movements of container trollies/cranes so that the sway of containers (which occurs when they are being moved) is reduced.
  • This system is currently being used in Chinese port terminals, including those located in Zhenjiang and Haikou.
  • The automated anti-sway system helps port authorities "process containers faster and more efficiently", which is traditionally "a costly, time-consuming, and potentially dangerous activity."

Situational Awareness Technology for Container Vessels

  • SeaMachines Robotics partnered with Maersk, a container logistics firm.
  • Together, these companies installed computer vision technologies in Maersk's new ice-class container ships.
  • The computer vision technology works with LiDAR and perception software which collectively serves as an AI-based situational awareness technology. This technology serves to upgrade and augment the vessel's operations.
  • The technology improves "at-sea situational awareness, object identification and tracking capabilities," such as monitoring the vessel's surroundings and identifying potential conflicts, which allows for safer and more efficient operations.
  • The technology lays the foundation for an autonomous collision avoidance system. Collision avoidance is becoming increasingly important as vessel traffic has significantly increased in many ports, making it "more complex to manage the regional vessel traffic, which can increase the risk of an accident."
  • Overall, SeaMachines says that this technology reduces operational costs by 40% and increases productivity by 200%.

Shipping Container Recognition Solution

  • Adani, the largest ports and logistics company in India, partnered with Nanonets.
  • Together, the companies implemented an automated shipping container code recognition system using optical character recognition technology (OCR).
  • The solution allowed Adani to fully automate the process of capturing vital shipping container information (e.g. container numbers, ISOs, seals, license plate numbers, container damage, and other hazards). This information is used to help identify cargo and track containers and is necessary to be recorded accurately to ensure regulatory compliance, determine whether a container can be accepted into the port, and identify the container's destination.
  • The recognition system provides a real-time container tracking and monitoring tool and reduced the need for 45 office workers stationed at 15 gates.
Part
04
of five
Part
04

Computer Vision: Top Companies Part 1

KoiReader Technologies and ABEJA are two top companies in computer vision and logistics that work with smart ports.

KoiReader Technologies

1. Description

  • Dallas-based KoiReader Technologies was the runner-up in the 2019 Smart Port Challenge.
  • The company has designed a system that can scan images, extract unstructured textual data from these images, and convert them into useful structured data. The system is a visual recognition system that leverages machine learning.
  • The system was developed to address difficulties in verifying declaration and financial documentation in the maritime sector, which is known to be a laborious process.
  • The system has been shown to enhance the accuracy of documenting logistics information for regulatory clearance, and lower risks linked to "misdeclaration of dangerous goods."
  • On LinkedIn, KoiReader Technologies lists computer vision and logistics as two of its specialties.

2. Revenue

  • The revenue of KoiReader Technologies is not available in the public domain.
  • The company has 11 to 50 employees, according to LinkedIn.

3. Funding Rounds

4. Services

  • The services of KoiReader Technologies are categorized into (a) Suite of Cognitive APIs, (b) Transportation, Logistics & Trade, (c) Artificial Intelligence, (d) Advanced Image Processing, and (e) Technology.
  • KoiReader Technologies offers cognitive APIs for services such as bill of lading processing, invoice processing, cognitive scanning, order digitization, document classification, video feed data capture, and claims data capture.
  • In the area of transportation, logistics, and trade, it offers services such as freighttech platform integration, claims application integration, purchase order matching, document management platform integration, and mobile app integration.
  • In the area of artificial intelligence, it offers computer vision, natural language processing, and natural language understanding services.
  • In the area of advanced image processing, it offers cognitive scanning, image upscaling, watermark removal, ink stamp removal, and noise or distortion removal.
  • In the area of technology, it offers optical barcode reader services, optical mark recognition, and intelligent process automation.

5. Competitive Edge

  • KoiReader Technologies's competitive edge lies in its pool of experts and its freemium pricing model.
  • The company's Free service is inclusive of basic API, throughput, dashboard, cloud deployment, and email support services, and is limited to at most 100 pages of data per month. The company's Enterprise service, on the other hand, involves tiered pricing and includes customization, bulk document processing, cloud deployment, and 24/7 dedicated support. This pricing model gives the company an advantage over companies that do not offer free services.
  • The company's founders are very much knowledgeable in artificial intelligence, logistics, image processing, transportation, advanced data science, supply chain, and algorithms. They have known each other for a long time and have been innovating for nearly 20 years now.

ABEJA

1. Description

2. Revenue

  • ZoomInfo estimates the annual revenue of ABEJA at $7.3 million.

3. Funding Rounds

4. Services

  • ABEJA categorizes its services into AI Consulting, ABEJA Platform, and ABEJA Insight.
  • ABEJA offers artificial intelligence consulting services to help businesses utilize, apply, and implement artificial intelligence. Examples of artificial intelligence consulting services that ABEJA has provided include behavioral analysis, product classification, demand prediction, and flow analysis.
  • ABEJA offers the ABEJA Platform to support the development, deployment, and operation of machine learning or deep learning.
  • ABEJA offers ABEJA Insight to enable businesses to perform retail store analysis via artificial intelligence.

5. Competitive Edge

  • The company's competitive advantage is its extensive experience in the area of artificial intelligence consulting.
  • It has supported the implementation of over 200 machine learning and deep learning projects and has worked with more than 150 companies.

Research Strategy

To identify top companies in computer vision and logistics that work with smart ports, we referred to PIER71's list of finalists for the latest global Smart Port Challenge. PIER71 is a joint initiative of the Maritime and Port Authority of Singapore (MPA) and the National University of Singapore (NUS) Enterprise that aims to promote innovation in the maritime sector. From this list of finalists, we picked KoiReader Technologies and ABEJA, the top-ranking companies that use computer vision and logistics. Since information about funding rounds was required, we assumed that startups are of interest.

We were able to gather most of the required details about these companies (i.e., description, funding rounds, services, and competitive edge) by examining the website and press coverage of these companies and by consulting Crunchbase. Only the revenue of ABEJA is publicly available. The revenue of KoiReader Technologies could not be located in the company website, the news articles that cover the company, and business information websites such as ZoomInfo, Hoovers, and Glassdoor. The revenue could not be triangulated either because the company's market share is also not publicly available. Since KoiReader Technologies is a privately held company, it is not obligated to disclose its revenue.
Part
05
of five
Part
05

Computer Vision: Top Companies Part 2

Two additional, top companies in computer vision and logistics that work with Smart Ports are CLARITECS and Ocean Freight Exchange.

CLARITECS

1. Company Overview

  • On its Linkedin page, CLARITECS states that it "is a maritime solutions company which leverages on a strong foundation of maritime expertise to develop digital applications for decision makers in shipping operations and commercial management."
  • CLARITECS is a top company because (1) it was the "second runner-up" in the PIER71 2018 Smart Port Challenge "out of a total of 122 submissions" in November 2018 and (2) in the second quarter of 2019, it was scheduled "to complete [its] BunkerMAESTRO’s base system and test-bed with 5 bunker operators representing 33% of Singapore’s bunker fleet."

2. Revenue

  • The company's estimated, annual revenue is $2 million.

3. Funding Rounds

  • That funding round was pre-Series A in the amount of $600,000 (USD) "from Schulte Group's venture capital arm, INNOPORT, and an anonymous private angel investor."

4. Services

  • CLARITECS' flagship service/product is BunkerMaestro, which is "an algorithm-based SaaS platform incorporating information services from MPA’s Maritime Data Hub and MarineTraffic to provide data-driven insights for bunker scheduling clarity."
  • The BunkerMaestro platform is able to "[s]ynchroni[ze] the entire bunker supply chain" and "improve[s] bunker tanker fleet deployment efficiency with integrated auto-scheduling."
  • BunkerMaestro is able to increase the efficiency of work by a minimum of 30%, raise the utilization rate for bunker tanks by a minimum of 15%, and decrease the amount of communication time required by 50%.
  • Another service/product offered by the company is called Auto-Profiling, which is "a system that provides diagnostics of Mass Flow Meter bunker delivery data to detect anomalies and generate reports for bunker deliveries."
  • CLARITECS also offers a service/product called Asset Managment, which enables customers to obtain "the fair market value of almost all tankers and bulk carriers in the world."

5. Competitive Advantage

  • CLARITECS' competitive advantage is that it created and offers "[t]he first platform on the market built for bunker operators to manage their entire fleet schedule, auto-schedule vessels to jobs, and manage fuels with IMO 2020 in consideration."

Ocean Freight Exchange

1. Company Overview

  • Ocean Frieght Exchange (OFE) is a tech startup based in Singapore that offers an "AI-driven marketplace for charterers, ship owners, and brokers in the dry bulk, tanker, and gas markets. OFE offers predictive analytics for supply and demand, vessel tracking, calculated vessel arrival times, comparing vessels and costings, and many others on one consolidated platform using AI and predictive analytics."
  • The company was established in 2015.
  • This link is to the company's website.
  • OFE is a top company because it was the winner of PIER71's 2018 Smart Port Challenge.

2. Revenue

  • The only revenue data we found about the company was that it earned $200,000 in revenue from its first data deal. We looked elsewhere for other revenue information (PitchBook, Crunchbase, the company's Linkedin page, and articles), but no additional revenue information was provided in any of those sources.

3. Funding Rounds

  • The company appears to have received investments through two funding rounds, one of which was a Seed round that five investors participated in.
  • In June 2018, it was reported that "Ocean Freight Exchange . . . raised $3.3m[illion] in over-subscribed seed funding earmarked for company expansion."
  • A total of eight investors have provided funding for the company.

4. Services

  • The company states that it offers the following services: Position list, vessel history, distance calculator, voyage calculator, fixtures, ports, bunkers, and trade flow.
  • The following features are offered collectively through the aforementioned, combined services: "Real-time email scraping [;] Live open tonnage list [;] Satellite and terrestrial AIS [;] Live open cargoes and orders [;] Owner and Charterer contacts [;] Trade flow [;] Voyage Calculator [;] Distance Calculator [;] Latest transaction fixture data [;] Port details, congestion and live traffic [; and l]ive bunker ports prices."
  • The aforementioned services and corresponding features are offered collectively as a package, which is available through monthly or yearly payment.
  • The monthly price for each user is $299.
  • The yearly price per user is $2,988 ($249 per month x 12 months).
  • The Enterprise package is available for 10 or more users and the cost is not specified.

5. Competitive Advantage

  • Per OFE's CEO, the company is "proud to have no conflict of interest, unlike others."

Research Strategy

To identify additional, top companies in computer vision and logistics that are working with Smart Ports, we looked for articles about such companies. From those searches, we found lists of winning and runner-up companies from the PIER71 2018 Smart Port Challenge in articles from Seatrade Maritime News and e27. Most of the information we found about the companies came from their websites, Linkedin pages, and articles that the companies were mentioned in. Revenue and funding round data mainly came from ZoomInfo, PitchBook, and CB Insights.
Sources
Sources