Deus ex Machina
Artificial intelligence and machine learning are unlocking the secrets of Big Data and helping transform maritime.
(Article originally published in Sept/Oct 2020 edition.)
In the Port of Singapore, a vessel is quarantined at anchorage due to COVID. A ship’s officer is connected to either a class surveyor or an IBM Watson-powered Artificial Intelligence (AI) agent via voice-driven, industrial-grade smart glasses. With smart glasses’ ability to feed audio, video and even interactive 3-D augmented- or mixed-reality (AR/MR) digital images over the natural environment, the surveyor/agent instructs the officer as to where and what to look for to credit survey items.
The evolution of human intelligence and technology has always been driven by data. Our species was only able to transition from hunter-gathering tribes to modern agrarian societies once celestial-derived knowledge was attained and new tools developed. With technology’s exponential integration into everyday life, when will computers’ promise of freeing us instead of further enslaving us begin? How do we make sense of the infinite amount of data currently generated?
As stated by the antagonist, Nathan, in the 2015 film Ex Machina, “Here’s the weird thing about search engines [Big Data] …it’s like, striking oil in a world that hadn’t invented internal combustion. Too much raw material, no one knew what to do with it.” Now, with AI, we have deus ex machina (“god from the machine”), and these companies are making sense of this raw material.
Singapore-based Internet of Things (IoT) provider Cerekon is offering intrinsically safe, industrial-grade smart glasses and manufacturing solutions utilizing computer vision (CV) and machine learning (ML) via IBM Watson’s AI architecture.
Founder & CEO Rohit Deshmukh explains how 85 to 90 percent of maritime companies still use paper-based or mobile tablets for inspections. From a safety aspect, hands-free wearable devices allow surveys to continue even while climbing ladders. CV provides guided on- or offline inspection with indoor navigation, virtual checklists, work instructions for maintenance and repair of equipment as well as capturing image and video for objective evidence.
“COVID has placed big pressure on maintenance and upkeep onboard vessels,” Deshmukh says. “Scheduled inspections have been deferred and must now meet requirements – particularly those in restricted areas. Today, an officer can wear their smart glasses and go to the inspection areas with a class surveyor remotely guiding them.”
As of mid-September, Cerekon has been conducting proof-of-concept for inspections and remote support with Thome Group, Teekay, BP and Adriatic Shipping.
For warehouse management and enterprise resource planning systems, real-time object recognition can assist logistics by locating products, reading barcodes, updating inventory, picking validation and determining availability. Once online, this subscription as a service (SaaS) syncs-up with the cloud for analytics and report generation. Deshmukh claims Cerekon’s compressed data allows live video-streaming at 150 kbps (typically requiring 500-600 kbps), which equals wider global coverage for remote assistance.
Classification is going through a major technical transformation. Thanks to COVID, adoption has accelerated AI-driven inspections, remote inspection techniques (RITs) and condition-based surveys. Kash Mahmood, Senior Vice President for Digital Solutions at leading class society ABS, says the core objectives of condition-based surveys are safety and increased performance.
The ABS Digital Asset Framework incorporates IoT-connected sensors, ML, AI and software to enable the shift from a calendar-based to a condition-based schedule. Digital asset twins are developed, then broken down into modules that are scored based on 160 data variations.
These digital twins address double categories for production, which focuses on design capital expenditures and simulation, and operations – how the asset’s operational expenditures are behaving versus their design elements. The goal is a completely managed lifecycle that provides a better understanding of assets from cradle to grave.
Risk-based modeling compiles operational profiles through historical data, load and damage exposure, original equipment manufacturers’ recommendations and health-monitoring techniques from sensors and smart equipment. By identifying risk and pinpointing failure modes, the results help enhance structures, increase machinery reliability and improve environmental compliance for discharge and emissions.
Steve Grotsky, ABS’s Director of Commercial Operations, explains that remote surveys are part of the present and future through a mix of onboard surveyors and remote inspection techniques (RITs) that comply with rules and conditions. ABS has been aggressively scaling up its cloud-based remote survey platform.
Clients can now go online to check which inspections qualify for remote surveys and even book through SMS or WhatsApp. With a 24-hour turnaround, owners and operators can use their iOS or Android devices to call a surveyor for guidance, take pictures and videos and upload objective evidence for crediting.
Marine and offshore corrosion is estimated to cost the industry $50-$80 billion per year. RITs and AI-driven inspections can improve asset monitoring across survey and maintenance strategies. To determine corrosion and coating breakdown, inspection data gathered remotely from drones, crawlers or ROVs are automatically assessed using ML.
ML objectively identifies anomalies, effectively defines proper corrosion ratings and determines timelines to rectify. To assist vessels on the go, AI can tap into a vessel’s AIS positioning to provide geo-fencing predictives of what to expect upon arrival in a port. Aggregation of data records (in the order of tens of millions) from vessel loadline surveys, maritime mobile service identities and port state inspection results provide automatic notifications and recommendations that hedge against general or vessel-specific detention items.
Deciphering Big Data
“Big Data has been around for more than 10 years,” says Cosmo King, CEO of Seattle-based tech firm ioCurrents. “Everyone got big data and didn’t know what to do with it. It became myopic and not the big picture. Now, AI and ML relieves data analytics and begins telling us what it all means through answer sets. Digitization gives you the data. Our platform tells a story. One can only check things so many times a day with a clipboard. AI can help while continuously analyzing and relieving some of the decision fatigue.”
ioCurrents MarineInsights™ analytics platform is a SaaS with a hardware component. Its mission is to help predict failures, optimize fuel and improve maintenance. It helps companies (particularly small to medium enterprises) become more competitive by helping them know when and where to look for problems and opportunities. King helps demystify AI by stating that it’s not the Terminator but rather a means for inputs/outputs and solutions.
ioCurrents starts by shipping a small box with Wi-Fi and two ethernet cables to plug in for a month of data. This builds a digital twin that establishes a normal baseline for its Automated Anomaly Detection (AAD) system. From there, it adds prescribed equipment thresholds (i.e., if OEM widget goes above x degrees or reaches a certain age, it will fail).
Examples: a 30-minute heads-up before a bearing failure, or five days advance notice on a generator outage due to a variance typically not alarmed on OEM panels, or the system detects anomalies when throttle, bearing temperatures and/or fuel pressure don’t all linearly increase.
Cosmo says that ioCurrents’ products are even changing crew behavior as the owner/operator is notified when any vessel in their fleet is idling too long and building carbon. Engines not under load are turned off, which reduces engine hours and maintenance, consumes less fuel and lowers emissions.
Supply Chain Management
To bolster and accelerate market velocity, Joseph Hudicka, Managing Director of Neurored, is leading the charge for the Seamless Shipper Experience. “If you don’t invest in yourself,” he says, “the other guy will crush you. Resiliency is what is needed through digital transformation and connectedness.”
As a trusted Salesforce Partner, Neurored harmonizes freight forwarders, traders and global shippers by tying together sourcing, logistics and sellers on a single cloud-based supply chain management and transportation management system software suite. Its Salesforce app allows users to track and trace vessels, vehicles and cargo containers using satellite AIS, IoT and integration with carriers.
Neurored leverages Salesforce Einstein AI’s growing network. Derived from data gathered on every user action, it integrates invoicing plus customer relationship management and is capable of real-time automated quotes with historical and predictive analytics. Atop of Salesforce’s proven and trusted platform and with clients like LafargeHolcim – not to mention playing a critical role in deploying personal protective equipment to front line workers – Neurored is prime time.
A Higher Power
For technology to truly become a value-add, it must perform several functions: Help alleviate workload, increase efficiencies, reduce risk and not overly displace our ability to provide. However, humanity now requires a higher (processing) power to adapt.
With our newfound ability to harness the power of a god from the machine, the immediate future appears bright for those willing to boldly go.
Former class surveyor and ISO auditor Sean Holt is a regular contributor to The Maritime Executive.
The opinions expressed herein are the author's and not necessarily those of The Maritime Executive.