In a Jan 2024 survey by Everest Group, 68% of CIOs pointed out budget concerns as a major hurdle in kickstarting or scaling their generative AI investments. Just like estimating costs for legacy software, getting the budget right is crucial for generative AI projects.Misjudging estimates can lead to significant time loss and complications with resource management.
Before diving in, it’s essential to ask: Is it worth making generative AI investments now, despite the risks and the ever-changing landscape, or should we wait?
Simple answer: Decide based on risk and the ease of implementation.It’sevident that generative AI is going to disrupt numerous industries. This technology isn’t just about doing things faster; it’s about opening new doors in product development, customer engagement, and internal operations. When we speak with tech leaders, they tell us about the number of use cases pitched by their teams. However, identifying the most promising generative AI idea to pursue can be a maze in itself.
This blog presents a practical approach to estimating the cost of generative AI projects.We’ll walk you through picking the right use cases, LLM providers, pricing models and calculations. The goal is to guide you through the GenAI journey from dream to reality.
Choosing Large Language Models (LLMs)
When selecting an LLM, the main concern is budget. LLMs can be quite expensive, so choosing one that fits your budget is essential. One factor to consider is the number of parameters in the LLM. Why does this matter? Well, the number of parameters provides an estimate of both the cost and the speed of the model’s performance. Generally, more parameters mean higher costs and slower processing times. However, it’s important to note that a model’s speed and performance are influenced by various factors beyond just the number of parameters. However, for this article’s purpose, consider that it provides a basic estimate of what a model can do.
Types of LLMs
There are three main types of LLMs: encoder-only, encoder-decoder, and decoder-only.
Encoder-only model: This model only uses an encoder, which takes in and classifies input text. It was primarily trained to predict missing or “masked” words within the text and for next sentence prediction.
Encoder-decoder model: These models first encode the input text (like encoder-only models) and then generate or decode a response based on the now encoded inputs. They can be used for text generation and comprehension tasks, making them useful for translation.
Decoder-only model: These models are used solely to generate the next word or token based on a given prompt. They are simpler to train and are best suited for text-generation tasks. Models like GPT, Mistral, and LLaMa fall into this category. Typically, if your project involves generating text, decoder-only models are your best bet.
Our implementation approach
At Robosoft, we’ve developed an approach to solving client problems. We carefully choose models tailored to the use case, considering users, their needs, and how to shape interactions. Then, we create a benchmark, including cost estimates. We compare four or five models, analyze the results, and select the top one or two that stand out. Afterward, we fine-tune the chosen model to match clients’ preferences. It’s a complex process, not simple math, but we use data to understand and solve the problem.
Where to start?
Start with smaller, low-risk projects that help your team learn or boost productivity. Generative AI relies heavily on good data quality and diversity. So, strengthen your data infrastructure by kicking off smaller projects now, ensuring readiness for bigger AI tasks later.
In a recent Gartner survey of over 2,500 executives, 38% reported that their primary goal for investing in generative AI is to enhance customer experience and retention. Following this, 26% aimed for revenue growth, 17% focused on cost optimization, and 7% prioritized business continuity.
Begin with these kinds of smaller projects. It will help you get your feet wet with generative AI while keeping risks low and setting you up for bigger things in the future.
Different methods of implementing GenAI
There are several methods for implementing GenAI, including RAG, Zero Shot, One Shot, and Fine Tuning. These are effective strategies that can be applied independently or combined to enhance LLM performance based on task specifics, data availability, and resources. Consider them as essential tools in your toolkit. Depending on the specific problem you’re tackling, you can select the most fitting method for the task at hand.
Zero shot and One shot: These are prompt engineering approaches. The zero-shot approach involves the model making predictions without prior examples or training on the specific task, suitable for simple, general tasks relying on pre-trained knowledge. One Shot involves the model learning from a single example or prompt before making predictions, which is ideal for tasks where a single example can significantly improve performance.
Fine tuning: This approach further trains the model on a specific dataset to adapt it to a particular task. It is necessary for complex tasks requiring domain-specific knowledge or high accuracy. Fine tuning incurs higher costs due to the need for additional computational power and training tokens.
RAG (Retrieval-Augmented Generation): RAG links LLMs with external knowledge sources, combining the retrieval of relevant documents or data with the model’s generation capabilities. This approach is ideal for tasks requiring up-to-date information or integration with large datasets. RAG implementation typically incurs higher costs due to the combined expenses of LLM usage, embedding models, vector databases, and compute power.
Key factors affecting generative AI investments(Annexure-1)
Human Resources: Costs associated with salaries for AI researchers, data scientists, engineers, and project managers.
Technology and Infrastructure: Expenses for hardware (GPUs, servers), software licensing, and cloud services.
Data: Costs for acquiring data, as well as storing and processing large datasets.
Development and Testing: Prototyping and testing expenses, including model development and validation.
Deployment: Integration costs for implementing AI solutions with existing systems and ongoing maintenance.
Indirect costs: Legal and compliance and marketing and sales.
LLM pricing
Once you choose the implementation method, you must decide LLM service (refer table 1 below) and then work on prompt engineering — that’s part of software engineering.
Commercial GenAI products work on a pay-as-you-go basis, but it’s tricky to predict their usage. When building new products and platforms, especially in the early stages of new technologies, it’s risky to rely on just one provider.
For example, if your app serves thousands of users every day, your cloud computing bill can skyrocket. Instead, we can achieve similar or better results using a mix of smaller, more efficient models at lower cost. We can train and fine-tune these models to perform specific tasks, which can be more cost-effective for niche applications.In the above table 1, “model accuracy”estimates are not included because they differ based on scenarios and cannot be quantified. Also note that the cost may vary. This is the current (as of July 2024) cost listed on the provider’s website.
Generative AI pricing based on the implementation scenario
Let’s consider typical pricing for the GPT-4 model for the below use cases.
Here are some assumptions:
We’re only dealing with English.
Each token is counted as 4 letters.
Input: $0.03 per 1,000 tokens
Output: $0.06 per 1,000 tokens
Use case calculations – Resume builder
When a candidate generates a resume using AI, the system collects basic information about work and qualifications, which equates to roughly 150 input tokens (about 30 lines of text). The output, including candidate details and work history, is typically around 300 tokens. This forms the basis for the input and output token calculations in the example below.
Retrieval Augmented Generation (RAG) is a powerful AI framework that integrates information retrieval with a foundational LLM to generate text. In the case of resume builder use case, RAG retrieves relevant data based on the latest information without the need for retraining or fine-tuning. By leveraging RAG, we can ensure the generated resumes are accurate and up-to-date, significantly enhancing the quality of responses.
Fine tuning cost
It involves adjusting a pre-trained AI model to better fit specific tasks or datasets, which requires additional computational power and training tokens, increasing overall costs. For example, if we fine-tune the Resume Builder model to better understand industry-specific terminology or unique resume formats, this process will demand more resources and time compared to using the base model. Therefore, we are not including the cost for this use case.
Summary of estimating generative AI cost
To calculate the actual cost, follow these steps:
Define use case: E.g. Resume builder
Check cost of LLM service: Refer to table 1.
Check RAG implementation cost: Refer table 3.
Combine costs: LLM service, RAG cost, and calculate additional costs (Annexure-1) such as hardware, software licensing, development and other services.
The rough estimate would be somewhere between $150,000 to $2,50,000. These are just the ballpark figures. The costs may vary depending on your needs, LLM service, location, and market condition. It’s advisable to talk to our GenAI experts for a precise estimate. Also, keep an eye on the prices of hardware and cloud services because they keep updating.
At Robosoft, we believe in data democratization—making information and data insights available to everyone in an organization, regardless of their technical skills. A recent survey shows that 32% of organizations already use generative AI for analytics. We’ve developed self-service business intelligence (BI) solutions and AI-based augmented analytics tools for big players in retail, healthcare, BFSI, Edtech, and media and entertainment. With generative AI, you can also lower data analytics costs by avoiding the need to train AI models from the ground up.
Image source: Gartner (How your Data & Analytics function using GenAI)
Conclusion
Generative AI investments aren’tjust about quick financial gains; they require a solid data foundation. Deploying generative AI with poor or biased data can lead to more than just inaccurate results. For instance, if a company uses biased data in its hiring process, say gender or race, it could discriminate against certain people. In a resume-builder scenario, this biased data might incorrectly label a user, damaging a company’s reputation, causing compliance issues, and raising concerns among investors.
While we write this article, a lot is changing. Our knowledge about generative AI and what it can do might differ. However, our intent of providing value to customers and driving change prevails.
‘Tis the season…the holidays are once again upon us, and with retailers prepping months in advance for the advent of the 2018 season, the goal as always remains: keep customers happy and keep them coming back for more. The currently crowded e-commerce landscape includes behemoths such as Amazon, Alibaba, and Walmart (to give perspective, Walmart is the largest retailer in the world, and Amazon, at number 7 on the list, had a market share of 37% in 2017), so smaller retailers are using every available tool to stay in the game.
As e-commerce continues to grow, with 2018 poised to be the biggest year yet, brick and mortar stores will be forced to use such technologies as artificial intelligence and machine learning (AI and ML) to keep their edge and outperform their competition, pursuing innovative and resource-conscious ways to connect with their digitally savvy consumer base.
To successfully compete in the 2018 holiday market, brand owners will need to ask:
What is it that consumers want from retailers this holiday season?
What functions will AI & ML influence this season?
What are the factors that brand owners should keep in mind while incorporating an AI strategy this holiday season?
In this article, we address those questions and examine how AI and ML technologies are forever shifting how customers and retailers engage and interact with one another, particularly during the frenzy of holiday shopping.
The Rise of Online Shopping
The e-commerce market in the United States is about $450 Billion per year, and e-commerce represents roughly 10 percent of the total market. On an average, consumers spend more than 3 hours per day on social media, and with the constantly evolving technological advances, search engines and websites aren’t the only platforms to see a dramatic rise in shopping activity. Social media sites such as Instagram have experienced a rise in their engagement. Consumers are more likely to respond and relate to direct messages sent by social media influencers than those sent directly from brands themselves. Brands such as Natori recognize the power of these organic touch points with potential customers, and the relationship-building groundwork they provide for future and repeat business.
According to the Deloitte 2017 Holiday Retail Survey, more than half of shoppers reported that they would make most of their purchases online for the 2018 holiday season. Past behaviour showed that consumers primarily use search engines to find the best deals and compare prices online, but the majority still went to a physical store to buy those products. Now, the scales have tipped, balanced between physical and online sales.
Succinctly put, consumers want more with less. They want an easy-to-use, issue-free shopping experience that promises to keep on giving, very much in the spirit of the season. They want the ease of price matching and to research customer reviews before they settle on which retailers will get their dollars, but they also want fewer choices when it comes time to actually making that purchase, to reduce the anxiety of the paradox of choice.
As soon as a customer decides to purchase a type of product, “having a smaller range of potential products to choose from reduces choice overload.” AI and ML play an important role not only in strengthening marketing efforts through retargeting ad campaigns but also, more importantly, in sending the right recommendations tailored for consumers (reducing anxiety and giving customers a sense of feeling taken care of) and providing consistently stellar customer service through voice assistants and personalized experiences.
There are several key areas where retailers are or should be harnessing AI and ML technology, in both brick and mortar and e-commerce, to increase their sales revenue this year for the holiday season.
Methods to the Holiday Madness
Christopher Schyma, the Director of Retail at Sutherland, is quoted as saying:
The rise of automation through AI will have a much more significant impact on retail than other industries . . . This is the result of the changing retail industry – today’s customer requires digital-first experiences, where needs are met and expectations exceeded across a variety of touch points and at the complete convenience of the shopper.
Retailers will try to woo consumers this year with:
Marketing: Retargeting campaigns:
Customer experience: AI-based recommendations
Supply chain & logistics: Product stocks and shipping
Customer service: Chatbots
Marketing
Retargeting campaigns are by no means a new player in the retail scene. However, AI technology is becoming more sophisticated, allowing you to glean more nuanced data from consumers’ online and purchasing habits.
Instead of simply suggesting that they buy the thing they saw online in one of your ads, retargeting campaigns using AI can suggest complementary products as well, or products that they may need in the future.
Customer Experience
Personalized shopping experiences, both in-store and online, are one of the key aspects that retailers refine and prioritize when crafting their overall brand experience. AI has been involving with this process in small ways for the past few years, but we’re seeing a greater adoption of technology-forward processes. Shoppers input their preferences into their online profile, and the system generates not only a customized profile for them to view but also a tailored shopping excursion, whether they choose to visit the store in person or shop online.
This level of customization saves the customer so much time sifting through a myriad of products to help them find just the right ones. “AI tech is getting so good that it knows what you want—and can suggest complementary products—even better than you do.”
Customer Service
Self-checkout tills and chatbots have become almost ubiquitous in most customers’ shopping experiences. Recent data from Accenture shows that 65% of consumers “are already using or would like to use a number of technologies that are powered by AI when shopping, including chatbots.” Retailers can use chatbots not only in customer service scenarios to cut down resolution times but also in the lead generation process.
The key for retailers will be to focus on freeing up their employees to do more complex, human-focused activities, rather than the mundane tasks that eat up time and can cause frustration and confusion for customers.
Supply Chain
Retailers can also make use of AI and ML technologies on the back end of their operations through their CRM, supply chain management, and logistics processes. Andrew Cross from Brabners says that AI will have a “serious impact on retail [through] the supply chain . . . inventory management systems are already in play.”
Companies such as Walmart have an AI system in place to solve their “last mile delivery issues” that frequently plague retailers when their stock is low or depleted. Richard Cawston, supply chain managing director for XPO Logistics, shares that Nestle has invested in “a custom-designed distribution centre that will feature integrated data analytics, intelligent machines, advanced sorting systems and robotics,” which will enable Nestle to streamline their operations in order to cut costs and better meet the demands of their customers over the holiday season.
Examples of Retailers Using AI Technology
Below are a few examples of retailers successfully implementing AI technology directly into their customer interactions, recreating experiences from the very first moment and giving customers a sense of awe and inspiration.
Tommy Hilfiger used chatbots heavily during New York Fashion Week in order to increase traffic to their site. The bots were nearly indistinguishable from a human, giving users the power to direct the conversation and providing them options on style guides and digital catalogues of their recent collection, as well as offering a “backstage” look of their latest show.
Stitch Fix makes use of sophisticated algorithms based on profiles filled out by customers on their likes and preferences. These profiles then help customers pick out the right items…not only saving the customer time and money but also ensuring your customers feel that each experience with your company has been tailored and personalized before they visit your online or brick and mortar store.
Japan’s SoftBank telecom operations created a humanoid robot ‘Pepper’ that could interact with customers and ‘’perceive human emotions’’. Pepper was deployed at the Ave apparel store and the store experienced a boost of 98% in customer interactions, a 20% increase in foot traffic and a 300% increase in revenue.
Both The North Face and the Mall of America use the IBM Watson-enabled platform to offer tailored shopping experiences and customized lists of items. The system, “through voice recognition technology queries and sentiments of customers,” can understand and create a personalized plan for each customer.
There are AI-enabled platforms that can help transform the entire purchase cycle from merchandizing to customer engagement for retailers, like Vue.ai. It is the world’s only end-to-end Artificial Intelligence stack for brands and retailers. Vue.ai’s product suite contains solutions for merchandizing and operations, the eCommerce site and app and all the way through to marketing.
Best Practices When Incorporating AI
Keep the human touch:
Retailers can embrace AI without losing the human values they built their businesses on. Your customers are people, as are your staff. Never let that become an afterthought.
Remember to be subtle:
Customers may feel “spied on” if retailers take too aggressive of an approach. Proceed with care, and remember that the choice to buy is up to the customer.
Privacy is paramount:
Your customers need to know that you have their best interests at heart. Put safeguards in place, no matter what technology you use, to ensure neither you nor your customers will regret sharing their information.
By using artificial intelligence in every aspect of your retail operations, from online searches to tailored suggestions to your supply chain management, you can ensure your customers feel valued and appreciated rather than feeling like a dollar sign. You can bring the human-ness back to holiday shopping…with a little (or a lot of) help from robots.
From IBM Watson winning Jeopardy less than a decade ago to Artificial Intelligence becoming a part of our daily lives through voice assistants like Siri, Google Home or Alexa, this technology has come a long way.
Recently in an interview, Google’s CEO, Sundar Pichai stated –
‘’AI is one of the most important things humanity is working on. It is more profound than, I dunno, electricity or fire.’’
While it might take some time for AI to become as ubiquitous as electricity in our lives, we are heading towards that direction. And, enterprises are betting high on the technology. According to a report, the AI market will grow at a rate of 52% by 2025. As enterprises boost their investments in AI, the reign of AI is just beginning to reshape and push innovations across industries like healthcare, manufacturing, retail, etc.
Here are some interesting innovations that we have seen in recent years revolutionizing various industries:
Healthcare
The AI healthcare market is expected to reach $6.6 Billion by 2020. The healthcare industry seems to be bullish about the technology and is using it in multiple ways.
Precision medicine is one discipline of healthcare where AI has proven to be extremely useful. Here a patient’s DNA is scanned through deep genomics algorithms, to identify anomalies that could be linked to genetic disorders and mutations linked to diseases like cancer.
Another prominent example of how AI is accelerating healthcare’s efforts in saving lives is Atomwise’s AI, which was able to predict two drugs that could put a stop to the Ebola virus epidemic. In less than one day, their virtual search was able to find two safe, already existing medicines that could be repurposed to fight the deadly virus.
Retail
Japan’s SoftBank telecom operations created a humanoid robot ‘Pepper’ that could interact with customers and ‘’perceive human emotions’’. According to Softbank Robotics America, a pilot of the Pepper in stores in both Palo Alto yielded a 70% increase in foot traffic in Palo Alto. Nestle used ‘Pepper’ to serve coffee at its stores in China and Japan. A visitor chooses the type, size and strength of coffee using the tablet held by Pepper. Once the selection is made, the humanoid passes the order to a dual-arm robot, which makes coffee with a Nestle coffee machine and places the beverage on the serving tray. The entire process takes exactly three minutes.
North Face an apparel brand has also adopted IBM Watson’s cognitive computing technology to help consumers with purchase decisions.
General Electric’s (GE) has created Predix Manufacturing Execution Software, which is designed to make the entire manufacturing process—from design to distribution and services—more efficient and hence save costs. This suite of solutions powered by data integration, the Industrial Internet of Things (IIoT), machine learning, and predictive analytics, provides manufacturers with plant-floor and plant-wide collaborative visibility of all work in process.
Banking and Financial sector
AI in the Banking and Financial sector can provide zero-lag customer service, improve efficiency and accuracy.
Commonwealth Bank of Australia (CBA) launched its in-house bot Ceba to more than a million customers. Swiss bank UBS last year launched its AI systems on the trading floor, which analyses the sea of market data to identify trading patterns and formulate new strategies for trading volatility for the bank’s clients.
AI is also helping in the compliance and security aspects of banking. HSBC partnered with big data startup Quantexa to utilise AI software to counter money laundering. According to Quantexa’s press release, “the technology will allow HSBC to spot potential money laundering activity by analysing internal, publicly available, and transactional data within a customer’s wider network”.
AI boosting efficiency in business operations
While AI has accelerated the pace of innovation across industries it has also permeated the foundations of business operations and is set to change the day to day work lives impacting the ROI of enterprises. A report from Accenture states that, by 2035, AI has the power to increase productivity by 40 percent or more, for enterprises.
Here is how enterprises are using AI and Machine learning in various functions for higher efficiency and boosting ROI
Marketing
AI can find multiple applications in the marketing function from search to customer engagement.
Here are few examples of how marketing is leveraging AI
Improved search functions – technologies like Elastisearch are becoming mainstream in e-commerce by populating the best possible results for a search query. The distributed nature of Elasticsearch enables it to process large volumes of data in parallel, quickly finding the best matches for consumers’ queries.
Customer segmentation is another aspect where AI is improving efficiency for marketing by learning from customer behaviour for e.g. Companies such as AgilOne are helping marketers to improve and optimize email and website communications, by analyzing, continually learning from user behavior.
Recently IBM launched its new IBM Watson AI Marketing Suite that improves the marketing efforts by personalized targeting, improved programmatic buying insightful campaign analytics. The suite contains three AI-based solutions – IBM Watson Ads Omni, IBM Media Optimizer, and Predictive Audiences.
At Robosoft, we were a part of a project aimed at helping a leading US retailer based in Illinois, draw maximum ROI out of their marketing spends. The company relied on senior managers’ making marketing investment decisions based on past experiences with traditional marketing channels such as billboard and newspaper advertising. Even though there were massive amounts of data available for analysis, no data was utilized in determining the best channels to spend marketing dollars for maximum return on investment.
As a solution, a large scale machine learning system was developed to maximize Marketing Return on Investment (MROI) across all digital marketing channels. The system helped in
Enhancing the company’s customer touch points’ data collection capabilities across all web & digital assets.
Determining customers’ purchase behavior across digital channels.
Recommending optimal allocation of marketing budget across all digital channels.
The company saw improved EROI across all digital channels, prediction model empowered marketing managers to make data-driven decisions and also helped in defining the in-depth content strategy to rank highly for the most relevant keywords.
HR and Recruitment functions
AI can help in various aspects of HR like scheduling meeting, filtering candidates, reducing attrition, enabling a faster recruitment process, etc.
HiringSolved is an AI-powered recruitment tool that enables diversity during selection.
Mya, an AI recruitment tool expedite the process of recruitment by providing quick responses to applicants about their application and other information related to it
IBM Watson is working towards building such a predictive model for companies, that can predict attrition patterns amongst employees.
A US-based recruitment startup wanted to revolutionize how recruiters hire using artificial intelligence. Through their proprietary machine learning algorithm, they wanted to reduce the time and effort required to fill a job position for companies. We were a part of a project for the client aimed at developing a robust machine learning algorithm to best match candidates to a job opening.
The matching algorithm that was created achieved high rate in matching candidates according to the job openings, making the recruitment process highly efficient.
Customer service and customer engagement
Artificial Intelligence is currently being deployed in customer service. According to Gartner, by 2020, 55% of all large enterprises will have deployed at least one bot or Chatbot.
Since chatbots can lead to faster but at times inefficient and machine-like customer responses, enterprises are using bots which can work in tandem with their human counterparts. One company that provides AI-augmented messaging is LivePerson, where simple questions can be handled directly by a bot, but as soon as the conversation becomes too complicated the bot can hand the conversation off to a human.
AI can also help in creating models to boost engagement with customers by improving internal processes. One of the largest pharmaceutical companies in Asia that conducts clinical trials based on various types of drugs that belong to Therapeutic Areas like Gastroenterology, Neurology, etc. wanted to ensure patients have access to a simplified explanation of the documentation given to patients during clinical trials. We were a part of a project for the client which was aimed at optimizing the process & time required in translation from Scientific to Simplified documents.
An AI based model was used to translate the documents to the desired language of choice. Additionally, a mobile app was built for the patients as an engagement platform for clinical trials which consisted of an AI-Chat bot that provides answers to user’s text/voice-based questions on-the-go. Resulting in patients being more willing to participate in the trial as they felt more in control of the clinical trial experience.
Supply Chain Management
One of the most challenging aspects of managing a supply chain is predicting future demands for production. Machine learning algorithms can find new patterns in supply chain data daily, without needing manual intervention or the definition of taxonomy to guide the analysis. Lennox International Inc. is an intercontinental provider of climate control products for the heating, ventilation, air conditioning, and refrigeration markets use machine learning for their demand forecasting.
AI can also help in automating the inspection process for the manufacturing enterprises for e.g. The machine learning algorithms in IBM’s Watson platform can determine if a shipping container and/or product were damaged, classify it by damage time, and recommend the best corrective action to repair the assets.
‘’The key to the digital transformation of accounting and financing is pairing people and machines together allowing each one to contribute in areas they are best skilled at. Machines can efficiently and accurately analyze a tremendous amount of data, they can spot patterns in the data and learn how to treat various kinds of data.’’.
Some organizations are using AI to simplify their finance and accounting process simple like-
At Deloitte, auditors access AI tools with natural language processing capabilities to interpret thousands of contracts or deeds.
At Crowe Horwath, data scientists have harnessed technology to tackle complex billing problems in the healthcare industry. The team used machine-based learning to sift through enormous but disparate billing systems of its healthcare clients to flag accounts with discrepancies.
Challenges in deploying AI to business process
Like with many emerging technologies, there are challenges, with deploying AI to enterprise processes. According to a new MIT-Boston Consulting Group survey, 85% of executives believe AI will change business, but only 20% of companies are using it in some way, and just 5% make extensive use of it. Some of the challenges that may impede the process are –
Access to data – companies need to invest in creating the infrastructure to collect and store the data they generate and to recruit talent capable of making use of it.
Ever changing markets – businesses do not work on a static model, which means AI models will decrease significantly in efficacy, so smart companies will need to keep deploying resources and investments in keeping up with the market dynamics.
Specialists – AI still being a niche domain, the lack of AI know-how in management is hindering its adoption in most cases.
Cost – AI technologies are an expensive deal to an organization. While big names have separate budget allocations for AI implementation, it is the small and mid-size enterprises that struggle to implement AI solutions to their business processes.
Computation Speed – Technologies like AI, machine learning and deep learning solution, require a huge number of calculations to be computed at hypersonic speed. This requires processors that have advanced processing power much higher than what is in general adoption today.
“In the end, all technology revolutions are propelled not just by discovery, but also by business and societal need. We pursue these new possibilities not because we can, but because we must.”
The bigger technology players are paving the way for having automated and AI enabled processes. The industry as a whole has to evolve in terms of technology, trained resources etc., the costs of deployment will need to go down for smaller players to be a part of the AI revolution and finally infrastructure at an optimal cost will need to be created.
Trends have been the hallmark of the fashion industry over the last few decades. Such trends are usually short lived – they probably last for a season or so. Now we live in a social media world, where trends last less than a day! But on a serious note, paying attention to game-changing trends in the mobile & tech world makes sense for enterprises lest they miss out on opportunities which impact their very existence. If Kodak had only paid attention to how photography changed through digital cameras and then mobile phones, the company would still be thriving in an era where selfies and point & shoot are common
When the first commercial flight took off over a century ago, carrying 1200 passengers in a short span of 4 months, little did we know that the airline industry is going to change the way people travel forever. Back then, it was a mode of travel only for the elite; the first air ticket ever sold was purchased for $400 (a sum that would be equal to $10,000 today). Today, air travel has become accessible to most and carries a humungous 3.7 billion passengers per year. The airline industry has not just seen an avalanche of passengers but also an immense shift in the way it operates, thanks to the new and exciting technologies impacting the airline landscape.
The industry was firmly on cruise control when suddenly it crash landed in 2020. As the COVID-19 pandemic took over the world, one of the major industry to get hit severely was the aviation industry. The profits for airline carriers were at all time low and many private carriers shutting shop due to heavy losses. The total cumulative losses for aviation industry between the period of 2020-2022 is estimated to be $201 Bn.
But thanks to the ever-changing dynamics of the industry owing to the technological revolution and the ever-evolving nature of consumer behavior, the airline industry has managed to recover well from the unforeseen challenges and is on a trajectory to regain its earlier status. There are however still prevalent challenges pre-pandemic that have become major concerns for all airline organizations.
Some of the challenges that the airline industry has continuously been facing even in the pre-pandemic era are:
#1 Operational efficiency
#2 Declining workforce
#3 Increasing CX while maintaining operational costs
#4 Increasing fuel prices
#5 Maintaining quality service in wake of high customer demands
#6 Volatility in certain geo markets
In our article: How technology is giving flight to customer experiences in the airline industry, we detailed the way some challenges mentioned above impact the airline industry and how technology is helping them address these. We have also pitched for a new pricing model – Right Price Model, for airline business which tackles the important CX initiatives while maintaining operational costs.
Taking the discussion further, in this article we will look at the emerging technologies which are digitally transforming the airline industry and paving the way for the future of a digitally transformed and customer-centric airline industry.
1. Blockchain Technology
A blockchain is a distributed database that maintains a shared list of data or records.
Blockchain technology started out as one of the revolutionary solutions in financial sector. But the recent years have shown that this technology has many uses across multiple industries, including airlines.
The operation of the airline industry has a lot of moving parts and requires data sharing at multiple touchpoints from booking to arrival and even after that amongst multiple players including airlines, online travel portals, airports, immigration so on and so forth. Any leakage of the data through the entire cycle has the potential to not only diminish user experience and affect revenue but jeopardize passenger safety and security as well.
Benefits of blockchain technology in airline industry:
Identity Management & Record Keeping – Blockchain technology enables a hassle-free remote identity management system for airlines as compared to the laborious current biometric system. It can also help in safely keeping records of critical information such as passenger details, baggage onboarding, flight path, tracking down a lost entity, etc.
Cross Integrations for Seamless Travel Experience – Blockchain technology can help airlines turn the miles offered to the flyers into a more valuable asset that customers can use beyond the flying. It can integrate with other aspects of travel industry like ticketing, loyalty programs and non-airline logistical industries like transportation and hotels to create a unified seamless experience for travelers.
Building a robust data security system – With so many passenger records to maintain it is an uphill task for the airlines to manage and maintain the sanity and privacy of this data. Blockchain technology with a security wrapper can help in creating a unique and secure way of sharing and managing this information through the use of authorized access requirements.
Airline maintenance – Airlines have to deal with a humungous amount of data including cumbersome databases and sometimes manual binders when it comes to keeping a log of maintenance records. Blockchain can help the industry to ensure that these records are virtually recorded. Ensuring that all information regarding the procurement of the parts of the aircraft, the maintenance history, the person who has handled the maintenance, modifications that were done, etc. are recorded in real-time, virtually, and is accessible to relevant people when the records need to be accessed. All these details available at the right time without the chances of any error can improve the practice of maintenance, security, and safety to new levels.
Examples of blockchain technology uses in the airline industry:
– Air France deployed blockchain technologies to create a COVID-19 test verification system via a mobile app during the pandemic.
– British Airways and Zamna are working towards facial recognition features using blockchain.
– Etihad Airways partnered with Winding Tree for their blockchain travel platform.
– Singapore Airlines uses blockchain technology for their frequent flyer loyalty program using KrisPay. It also offers promotions to customers along with the program.
2. Augmented Reality and Virtual Reality
Industries like retail, healthcare, etc. are seeing a lot of uses of the AR/VR revolution. The airline industry is also following suit. Right now one of the most obvious applications of these technologies can be expected to be seen in the airport arena. AI can not only lessen the burden on human resource but increase satisfaction when used for customer service in airports.
Examples of AR/VR uses by some airports to delight their customers:
– The Gatwick airport uses AR to help passengers navigate the complex layout of the airport, and London City Airport has installed AR tech to help air traffic controllers with the vital job of keeping planes safe.
– Vodka brand Grey Goose created a multi-sensorial interactive AR experience at Heathrow Airport to engage with customers. As part of this experience, the customers can unveil three new flavors of the brand with simple hand gestures.
– Istanbul Airport created a virtual character called CiGA through AR technology that’ll accompany passengers around retail stores. The character presents special offers from participating retailers and directs them within iGA Istanbul Airport International Arrivals Bosphorus area.
– Another example of AR technology enhancing the airport experience is San Jose International Airport partnering with the Google Tango team for a trial of augmented reality technology in wayfinding, through airport retail promotions and even AR billboards displaying destination information.
According to Jonathan Vaden, lead of the project “Once augmented reality technology becomes ubiquitous, we will see many new and creative ways in which airports around the world begin taking advantage of its capabilities.’’
– Airlines are also starting to use AR/VR technology to create delightful customer experiences for their customers. Qantas is one such airline that has launched a VR app that provides its passengers with immersive, engaging, and experiential videos of the beautiful Australian destinations.
– Together, American and Microsoft are applying the power of AI, machine learning and data analytics to reduce the taxi time for flights, giving connecting customers extra time to make their next flight while also saving thousands of gallons of jet fuel and decreasing CO2 emissions for the American Airlines fleet. Built on Azure, American’s intelligent gating program provides real-time analysis of data points, including routing and runway information to automatically assign the nearest available gate to arriving aircraft.
3. Artificial Intelligence
AI is gaining traction in present times and becoming more and more equipped to understand human interactions. Many industries are using it to upgrade customer experience at every touchpoint. From chatbots to voice-based AI tools there are umpteen use cases of AI being utilized.
AI integrated with machine learning, and predictive analytics can help immensely in providing a connected and customized experience to the flyers. Further, AI also has the potential to ease out various operational processes of airlines like revenue management, managing ticket pricing, etc. A lot of forward-thinking airlines understand the impact AI can have in multiple areas of the industry and are already investing in the same.
Examples of AI technology uses in the airline industry:
– Shenzhen airport in China uses AI for AI airbridge allocation as well as for AI turnaround times. Another application of AI in airports can be found in autonomous airside operations.
– UK-based EasyJet uses AI predictive analysis to make sense of all the available customer data and use these insights to create offers and services personalized for individual travelers. The airline also has a recognition tool that reads passports and fills out all the information for flyers—easing the data entry and data management tasks more manageable.
– Air France implemented the specialized AI platform called Sky Breath that collects data from the flight, performs in-depth analytics, and helps identify fuel-saving opportunities and increase efficiency.
– Delta Airlines installed four self-service kiosks at the St. Paul International Airport that used facial recognition technology to verify customer identity by matching customer faces to their passport photos.
4. Beacons technology
Beacon technology has seen a lot of success when it comes to retail. Given the fact that GPS or Global Positioning System is dependent on the low ranging indoor satellite signals, there is a huge potential for the airline industry to use Beacons in making navigation easy for travelers between different terminals at the airport. Further, Beacons can help airports and vendors at the airport premises to know where passengers are and then send them personalized and relevant information accordingly. These updates can be about boarding gate number, baggage carousel, flight status, or also about the shops and eateries around the customer.
Examples of Beacons technology uses in the airline industry:
– MIAMI International airport is already leveraging Beacons on its premises to create a personalized experience for travelers. The app provides information about the entire airport as travelers navigate through various places on the premises. Further, they are also updated with relevant information depending on their individual journey, e.g. gate numbers, flight updates, baggage collection details, etc. Additionally, with the ‘blue-dot’ functionality, map rotation, turn-by-turn directions, ‘walk times’, and a ‘near me’ feature, they allow passengers to quickly locate virtually anything inside the airport.
– Central America’s one of the largest regional commercial operators, Aeromexico Connect, are trialing Beacon platform for key maintenance bases for a few months. The beacon will be assisting all maintenance operations and accelerate an aircraft’s return-to-service time. Further, it will integrate the platform into the airline’s stations to include their on-call maintenance providers as well.
When we talk about use of robotics in aviation, we normally refer to manufacturing process of an airplane. As you know robotics help in various tasks during manufacturing such as drilling and fastening, welding, sealing and dispensing, inspection, rigid manufacturing, and transportation of parts.
However, the airline industry has also been using robotics in assisting various manual tasks at the airports. These include baggage handling, car parking, assisting with passenger check-in, temperature testing, or security, etc. Now, the implementation of robotics has gone beyond previous mentions as a customer-facing technology at airport terminals.
Examples of usage of robotics in aviation:
– The introduction of KLM’s socially aware ‘Spencer Robot’ in 2016 created a lot of buzz. This robot has been equipped with the capability to deal with social situations between people and can ‘see’ and analyze people nearby with his sensors. Spencer can also distinguish between individuals, families, and larger groups, and also learns about and then complies with social rules, ultimately acting in a human-friendly way.
– Airports like Glasgow Airport and airlines like Japan Airlines, EVA Air, etc. use robots as customer-service agents.
– Kansai Airport in Japan has deployed two security robots since 25 October 2021. These robots autonomously navigate and patrol routes, use a laser sensor to identify their locations and capture images with built-in cameras. They are programmed to stop and stand guard at their designated positions once their other tasks are finished.
– The robot named Pepper, launched by EVA Air, scans boarding passes to provide departure details, gives passengers weather updates for their destination, and informs them of duty-free special offers and airline promotions. Pepper can also play games, shake hands, dance, and pose for photos with travelers.
Biometrics is not new to aviation. All the major and minor airports started implementing it since 9/11 to improve their security details. But over the years it has found use in improving passenger experience as well by improving the time and speed of check-in and other operations.
Adopting Biometrics Technology at airlines and airport touchpoints is an attempt by the industry to add value to customers by delivering a delightful experience. For e.g. Delta Airlines launched the world’s first self-service biometric-enabled baggage drop to “free up more Delta people” to deal with customers.
The advantage of using biometrics at bag drop is two-fold: it helps in saving time for passengers and creating a visual record of the actual passenger dropping the bag – not just someone who might have picked up the suitcase owner’s boarding card.
Some examples of Biometric technology use in airports:
– Fraport in conjunction with Zwipe have agreed to trial their biometric solutions to boost security at Franfurt airport. The biometric data will not be stored in any centralized location which decreases the chance of a large security leak. Also, the use of contactless solutions decreases the chance of any spreading of germs or viruses in wake of the pandemic.
– Miami International Airport and US Customs and Border Protection (CBP) started rolling out biometric technology with a few airlines back in 2019. MIA is now seeking a huge biometric push by 2023 that will serve multiple purposes. Quoting Ralph Cutié, MIA Director and CEO here:
“We look forward to elevating our passenger experience with this state-of-the-art boarding solution. MIA is now the busiest U.S. airport for international travel and continues to set new records each month for passenger growth. Biometric boarding is one of the major steps we are taking to pave the way for additional growth in the years to come.”
Some of the critical locations where biometrics can be used are
Check-in
Bag-drop
Security screening
Duty free
Airline lounge
Border control
Self-service boarding gate
Destination immigration
Wearable technology
The wearable devices market was valued at $14.6 billion in 2019 and is expected to reach $85.6 billion by 2027, growing at a CAGR of 24.8%.
Initially designed to supplement the healthcare and fitness industry, wearables have found usage across other industries like retail, banking, and insurance to manufacturing and travel as well.
Airlines have started to use wearable technology in various ways to do more than improve customer experience on flights.
Some of the examples of airlines using wearable technology are:
– Hamad International Airport (HIA) has implemented the use of robotics and advanced thermal helmets in the post COVID-19 era as part of new measures being taken.
The Smart Screening Helmet is a wearable intelligent helmet, which is portable, safe and effective, and enables contactless temperature measurement.
– Japan Airlines use Microsoft’s HoloLens for training its new crew members and engineers. Using HoloLens, the mechanics can be trained in engine mechanics akin to the experience they will have working on an actual plane.
Microsoft has also demonstrated earlier that HoloLens can help in designing airport terminals by providing designers with imagery of the new terminals even before starting the construction.
– EasyJet and British Airways are among the airlines that have created apps for the Apple Watch, enabling passengers to store boarding passes and receive real-time updates on their wrists.
EasyJet has also introduced wearable uniforms for its crew and ground staff. The suit is fitted with hems and LEDs on the shoulders and can provide visual guidance to passengers. It is also equipped with built-in microphones to enable direct communication with passengers. Additionally, the uniform also has a LED-based scrolling ticker, on the lapel jacket, which displays basic information such as the number of the flight, the flight destinations, and lighting guidance in case of emergencies.
– Another, example of wearable devices being used to improve flight experience is the SkyZen app launched by The International Air Transport Association. The app is connected to the Jawbone fitness wristband and enables flyers to view their activity and sleep patterns throughout the flight and creates personalized insights based on their sleep and activity patterns.
8. Internet of Things
The objective of industries adopting IoT is to create a seamless and integrated ecosystem connecting organizational functions with the end consumers. The airline industry plunged into the IoT revolution with the same agenda – building a seamless and integrated ecosystem integrating the organizational functions to increase efficiencies and provide a seamless experience to their customers.
According to the FTE:
‘’Over the course of the next decade, it is likely that all “things” onboard will be connected and the health of everything, from engine performance to the IFE system, will be monitored in real-time. Sensors will automatically detect and report faults to maintenance teams on the ground, removing the need for the crew to manually report faults. Moreover, the addition of sensors to aircraft seats will enable the crew to monitor individual passenger health and wellbeing, and to proactively respond to their needs.’’
Examples of IoT in aviation industry:
– Virgin Airlines have implemented IoT in its Boeing 787. Every single element on the plane is attached to a wireless airplane network, providing real-time IoT data on elements like performance, maintenance, etc.
The airline is using the deluge of data that it is collecting through these flights to improve the efficiency of the aircraft and also be proactive. For example, a jet engine that is performing poorly mid-flight is relaying that information to ground staff throughout its journey, and when the plane lands, airport engineers can then be ready to look into the issue.
– Another example of the implementation of IoT by airlines is EasyJet’s Mobile Host at London’s Gatwick Airport. In this pilot, travelers’ flight details are combined with the live data from the airport’s Google indoor maps. This allows the airline to deliver updated check-in reminders, gate updates, and even personalized directions.
Though the implementation of a 360-degree implementation of IoT systems by airlines still has a long way to go, the process has definitely started on the right note.
9. Big Data
In a digitally connected universe, consumers leave their digital footprints at every touchpoint. Airlines can drive valuable insights by analyzing this data to create delightful experiences for travelers. E.g. Airlines can use this data to understand customer preferences in real-time based on data of their purchase history, travel itineraries, etc., and provide them with customized offers.
Some examples of Big data usage in aviation industry:
– United Airlines uses a smart ‘’collect, detect, act’’ system to analyze around 150 variables in the customer profile including their previous purchases, preferences, etc., and provide tailor-made offers to them. United Airlines has seen a YoY revenue increase of 15% after the implementation of this system.
Further, this data can also help in increasing operational efficiencies through predictive analytics. Predictive analytics can pre-empt any delays that might happen due to the weather forecast and in turn, inform the airline staff about it to keep their customers updated. Further, during any flight, a huge amount of data is generated with regards to pilot reports, incident reports, control positions, warning reports, etc. This data can be used in improving flight efficiency and safety.
– Southwest Airlines has partnered with NASA to indicate potential safety issues. By using machine-learning algorithms, they have built an automated system capable of crunching vast data sets to warn about anomalies and prevent potential accidents.
10. Mobile solutions
Today, smartphones have become an integral part of people’s life. With people spending a huge amount of time interacting with their devices, it has become increasingly important for industries to connect with their customers through the mobile platform.
Airlines have started venturing into the world of mobile solutions and are using this platform to connect with their customers throughout the passenger journey starting from booking a flight to deplaning it.
Examples of mobile solutions implementation in aviation:
– Delta Airlines recently started providing their passengers virtual boarding passes 24 hours before their journey through their mobile app, easing out the check-in process for their passengers.
Airlines not just are using mobile apps to make the entire process of booking flights, check-ins, etc. easier for travelers, they are also using apps in innovative ways to provide awesome in-flight experiences to the customers.
In our article, we take a detailed look at how Airlines are using the mobile platform to create delightful customer experiences.
5 ways how technology solutions can transform your Airport Facility Management ready for any pandemic level contingency
#1 IoT connectivity across user touchpoints
IoT enables the integration of data, systems, IoT connected tools, and mobile devices to create a unified platform to operate on. It detects data patterns left behind by customers and then translate the data into actionable insights. The insights act as recommendations and help detect problems before they occur.
IoT can be applied across various touchpoints of passenger journey from parking and departure to arrival and baggage claim.
Airport restrooms are using IoT equipped flush and urinal valves, faucet alerts for leaks and clogs, water usage monitoring and vacancy indicator lights – all these aid combine together to provide a streamlined experience for travelers.
Other aspects of airport operations, like retail revenue, luggage tracking and efficient terminal navigation can all be enhanced by IoT software.
#2 Touchless facility offerings
The coronavirus pandemic has been one such unprecedented event which forever changed our perception towards cleanliness and hygiene. The multitude of passengers coming and going from airports make them a prime hub for bacteria and pathogen transmission. Hence, the impetus is on airports to provide a touchless service across the whole journey of any passenger. According to Airport Council International, 89% of airports plan to implement touchless check-in self-service initiatives by 2023.
Major airports have already implemented touchless equipment upgrades across their facilities for both passengers and their employees. These equipment laced with features like biometric facial recognition and thermal scanning operations are making the airways a much safer way to travel.
#3 Smart restrooms across the entire arena
Restroom experience play a major role than anticipated in customer satisfaction when it comes to airports. Restrooms also act as an ideal place to gather real time data around restroom cleanliness and overall experience. Feedback devices placed around restrooms can notify facility management regarding immediate action required and save untimely cleanliness issues.
Devices and apps like autonomous floor scrubbers, attendant communication apps and automatic mobile alerts for faulty equipment help in providing a cleaner restroom experience. Also, some smart restroom apps can direct crew members to specific gates and times to respond to a specific day’s air traffic.
#4 Intelligent platform for airport operations
Airports can combine cutting edge technologies like AI, beacons, and mobile devices to optimize and streamline passenger experience.
Beacon technology can be used to help deliver real-time results like work order statuses, team member locations, audit results and other operational data. AI combined with beacons can provide real-time proximity data to mobile apps and drive airport cleaning efficiency, automatically dispatching the nearest team member to support emergent work orders.
Adopting up to date innovations and big data creates opportunities for airports to elevate passenger experience while optimizing operational efficiency, cost savings and resource management.
#5 Centralized control center
The fast transmission of coronavirus across international borders through air passengers have showed the importance of having a centralized data point and information center. Airports now find a routinely collected information about arrivals, departures and passenger counts more important than ever.
Airports can build a centralized dashboard and control center to have a single data source collected in one place to help airport teams make informed, long term, less reactive decisions. It can also help in having ultra-efficient labor management, data driven metrics to determine the macro effects of facility services and integrated predictive analytics. Also, a centralized data center can assist in proactive trend identification like passenger forecasts and seasonal peak models. These in turn help build proactively predictive models in case of potential surge.
In conclusion:
The course of travel industry was changed when Wright Brothers introduced the world to the flying machine. Today a century later the airline industry has grown by leaps and bound. Digital technologies are changing the landscape of every industry and the digital revolution in the airline industry has just begun. In times to come the airline travel experiences are set to become more personalized, valuable and memorable for the flyers.
The lines between digital, mobile, technology and apps are blurring in today’s world. We at Robosoft are keen followers of developments in this arena. Here’s a quick summary of what’s buzzing:
The announcements at Google I/O 2017, widely covered in tech media, are sure to have a wider impact beyond just mobile. Sundar Pichai, CEO of Google said in his keynote that ‘we are moving from a mobile-first to an AI-first world and we are rethinking all our products from that perspective’.
A few decades ago all we could do from our mobile phone was make a call. Later, the mobile phones evolved and started coming with simple Java games, calculator, text messaging and calendar. And now we have come to the era of smartphone apps that have dramatically affected the way we interact with our smartphones today. Soon, the mobile apps we know as of now would evolve further, changing the way we think, work and live our life.
The lines between digital, mobile, technology and apps are blurring in today’s world. We at Robosoft are keen followers of developments in this arena. Here’s a quick summary of what’s buzzing:
How healthcare companies can benefit
from the digital age.
Digital healthcare services can support a future where effective medical facility is available to everyone at reasonable cost. Find out how we can help you stay ahead with quality healthcare services....
With Facebook, Samsung, and Google getting serious about virtual reality, 2017 is expected to be a turning point for the VR industry.Content creators, hardware manufacturers and service or solution providers...
The experts guide to developing engaging
entertainment apps.
Communitainment apps are a growing rage with the millennials. Developing at least half a dozen of these video-on-demand apps like Voot, Colors TV, dittoTV and more for well-known media...
Essential guide for non-profits on mobile
app development.
Mobile apps can be a powerful marketing tool for non profits that focus on social good. A native mobile app can help drive awareness for the cause, encourage donation...
The secret to building great mobile apps: 3T's.
Download free e-book.
Among millions of mobile apps, only a few have been able to enjoy an incessant amount of engagement over the years. Building and sustaining a mobile app that constantly keeps users engaged...
Begin with these kinds of smaller projects. It will help you get your feet wet with generative AI while keeping risks low and setting you up for bigger things in the future. 
In the above table 1, “model accuracy” estimates are not included because they differ based on scenarios and cannot be quantified. Also note that the cost may vary. This is the current (as of July 2024) cost listed on the provider’s website. 
