Рубрика: AI

There’s effectively no area of modern business in which data science and analytics do not have some form of influence, and logistics is an example where they matter more than in most contexts.

In particular, it’s the application of these overlapping specialisms to the process of route planning that holds the most relevance for organizations operating today, so here’s a look at what aspects are most in the spotlight right now, and why they matter.

Advanced machine learning algorithms in route optimization

It’s no surprise that we’re seeing the widespread implementation of machine learning (ML) in route planning, thanks to its ability to offer up more precise and efficient solutions than were achievable via manually programmed methods.

Key innovations include:

• Reinforcement learning: Utilizes trial-and-error methods for continuous improvement in route decisions.
• Neural networks: Process complex datasets like historical traffic flows, weather conditions, and road closures to find the best routes.
• Clustering techniques: Group similar routes together for efficiency gains in delivery or transit systems.

Practical applications

This tech is already being put to work by all sorts of businesses, for the obvious reason that it both improves efficiency and saves money – to the tune of $50 million annually in some cases. Its abilities factor in the likes of:

1. Dynamic Routing: Adjusts paths on the fly based on current traffic situations.
2. Predictive Maintenance: Identifies potential vehicle failures before they happen, minimizing downtime.
3. Fuel Efficiency Optimization: Select routes that save fuel costs by avoiding congested areas or steep inclines.

These are the kinds of benefits that can of course be achieved by hand using widely available tools. For instance, you can optimize your route with Google Maps by adding stops and picking particular routes based on local knowledge. However, ML makes it swifter, slicker, and more business-focused.

Integrating IoT and sensor data into logistics planning

The integration of Internet of Things (IoT) devices and sensor data has turned logistics planning on its head. Collecting real-time information from various sources means companies can make informed decisions to optimize their routes and operations.

The key benefits here include:

• Real-time tracking: Monitors vehicle locations, traffic conditions, and delivery statuses continuously.
• Environmental monitoring: Sensors detect weather changes, road hazards, or unexpected delays.
• Vehicle diagnostics: Collects data on engine performance, fuel levels, and maintenance needs. Since the cost of fleet disruption due to vehicle downtime can spiral into thousands of dollars for a single day, this has far-reaching financial consequences.

Practical implementations

In terms of actually putting IoT sensors and the data they generate to work, there are a few main examples you need to know about:

1. Smart fleet management: Uses GPS tracking combined with sensor data to optimize fleet routes dynamically.
2. Condition-based maintenance: Schedules repairs based on actual wear-and-tear instead of regular intervals.
3. Enhanced safety protocols: Alerts drivers about hazardous conditions in real-time to prevent accidents.

These advancements enable logistics managers to respond proactively rather than reactively. It also applies across the supply chain, such as in the case of FedEx using IoT hardware and software to provide comprehensive visibility into its operations for customers in industries where this matters most, such as healthcare.

Autonomous vehicles and AI-driven routing decisions

Autonomous vehicles (AVs) combined with artificial intelligence (AI) are another player in the game of revamped route planning that’s ongoing worldwide as we speak. These technologies promise unprecedented efficiency, safety, and adaptability in logistics.

Some of the key advantages here are:

• Precision navigation: AVs use AI to navigate complex routes accurately.
• Continuous operation: Operate around the clock without fatigue or human error.
• Adaptive learning: As mentioned in regard to ML earlier, AI systems can also improve their decision-making over time based on accumulated data.

Practical applications

We’re still a long way from seeing AVs being the dominant force in global transport, and the market is only worth around $1.9 billion at the moment, but the future looks bright for the following reasons:

1. Last-mile delivery automation: Self-driving delivery vans optimize the final leg of deliveries, reducing costs and increasing speed.
2. Real-time route adjustments: AVs receive live updates on traffic conditions, road closures, or accidents to choose optimal paths instantly.
3. Platooning techniques: Multiple AVs travel in close proximity to reduce air drag and save fuel.

Once the kinks in this technology have been ironed out, it should not be long before the dream of truly autonomous delivery becomes a reality, and companies along with their customers reap the benefits.

Wrapping up

In short, route planning is an example of the overarching influence of data science and business analytics on commercial operations in the 21st century.

Businesses have more data than they know what to do with, so it makes sense to squeeze value from it in novel ways, whether that’s machine learning applied to route optimization or IoT for vehicle monitoring and maintenance.

IBM and The All England Lawn Tennis Club (AELTC) are collaborating to offer fans a more immersive and dynamic Wimbledon digital experience through a new feature called “Catch Me Up”. AELTC owns and operates Wimbledon, one of the most prestigious and historic tennis tournaments in the world.

Leveraging match data with generative AI capabilities from IBM's AI and data platform, watsonx, the new feature offers AI-generated summaries for every singles player, describing what happened in their last match and the challenges they face in their next one. It’s a quick and easy way to get a summary of the player’s progress through the championships, scheduled to start on July 1. The feature can also create long-form summaries for the day’s action.

Catch Me Up is available via the Wimbledon website and the Wimbledon 2024 App. Users can personalize the feature to add their favorite players to the list. The pre-match content offered by the new feature includes analysis of recent performance and likelihood-to-win predictions. Post-match content includes match highlights and key statistics. The feature also provides relevant links to any available Wimbledon live streaming or video highlights.

IBM has been at the forefront of integrating technology into tennis for several decades, enhancing the way fans and players experience the sport. The IBM Slamtracker, introduced in 2012, offers real-time match statistics and predictive analytics.

The advancements in AI have allowed it to offer more nuanced data analytics. In 2021, IBM enhanced its digital offering with new IBM Power Rankings and also introduced Match Insights with Watson.

“For 35 years, IBM and Wimbledon have been co-creating solutions that make fans feel more connected to all the on-court action, and our new research confirms they are beginning to understand the positive impact technologies such as generative AI can have on their digital experiences,” said Jonathan Adashek, Senior Vice President of Marketing and Communications for IBM.

The Catch Me Up feature is built using IBM’s Granite LLM to provide AI-generated text using the watsonx platform. The model has been trained on the Wimbledon editorial style to deliver a similar style and vocabulary. The AELTC and IBM editorial teams will monitor the outputs, providing a human backstop for potential corrections.

The new feature could potentially be most useful during the early stages of the championship, when there are multiple matches each day, including some that take place simultaneously. The launch of the new feature will also provide more coverage for lesser-known players, especially those competing in the juniors or wheelchair events.

The launch of Catch Me Up comes as new research from IBM and Morning Consult shows that more than 55% of global tennis fans surveyed believe that AI will have a positive impact on sports. The respondents expect the biggest improvement in real-time updates, personalized content, and unique insights.

Nearly one-third (31%) of the respondents use multiple devices to get more information, watch multiple matches at the same time, and interact with other fans. About half (47%) of the fans engage in additional content for news and rankings. Catch Me Up can offer a comprehensive solution by providing real-time updates, in-depth match analysis, and interactive features to meet the growing demand for a more engaging viewing experience.

The multi-decade partnership between Wimbledon and IBM sets the benchmark for technology integration in sports. The launch of the Catch Me Up feature is a testament to GenAI’s growing maturity and versatility. It shows how technology can add real value to the viewing experience for sports fans. Other sports organizations should take note as they explore AI applications.

Related Items

AI Bringing New Insights to the 2020 Summer Olympic Games in Tokyo

The Masters: A Tournament of Firsts Embraces the Future with IBM watsonx

Ferrari Revs Up the Performance of Its Sports Cars and Racing Team with AWS Cloud Services

In recent years, there has been an alarming increase in data breaches. In The first quarter of 2023 alone, over 6.4 million data records were breached worldwide. The increase in data breaches highlights the urgent need for both organizations and individuals to instill proper security measures. Failing to have adequate data protection measures invites a host of hackers into your system. No one wants to be a victim of such breaches.

Data breaches can be so devastating. They have serious financial and reputational repercussions. According to a recent IBM report, the average cost of a data breach was recorded at USD 4.45 million, which is a huge figure. It tells you of the importance of remaining secure at all times.

However, the big question is, what measures can you put in place to protect your online data? This article seeks to answer this question. It points out the top data protection tips that will keep you safe from the jaws of hackers. First, we need to understand what data protection is.

What is data protection?

Data protection includes practices, technologies, and regulatory measures designed to protect computer networks and resources against hackers. Data protection will secure personal and sensitive information from unauthorized access and loss.

At its core, the goal of data protection is to ensure the privacy and integrity of data by ensuring its confidentiality, security, and availability. To do so, data protection employs a host of measures such as encryption, access controls, data masking, and many others.

Tips to keep your personal information online safe

There are several effective tips you can adopt to keep your online data safe. This section spells out 10 practical tips you can apply to safeguard your data.

1. Use an SSL certificate to encrypt data

An SSL certificate is one of the most crucial elements for online data safety. If you run a website that collects or transacts sensitive data, consider having an SSL certificate. The certificate works by initiating encrypted sessions between the server and the browser. Through encrypted sessions, all information exchanged is unreadable to third parties.

The certificate prevents attackers from intercepting data while in transit. It, therefore, provides a secure environment for online transactions, data exchanges, and the safety of login credentials and other sensitive data. For a website owner, having a certificate can be a big plus. Remember, the certificate does more than just data encryption. It also helps with SEO, boosts user trust, and increases website performance.

2. Perform regular data backups

Performing regular backups safeguards your data against loss, corruption, and breaches. By routinely copying your data files and storing them in an alternative source, you ensure that you can easily recover the data in case of uncertainties. Uncertainty, in this case, might include data breaches, hardware and software failures, accidental deletions, ransomware attacks, or other forms of malware. While at it, please ensure you choose safe data storage locations, for instance. You can work with cloud storage systems, external hard drives, or both.

You should schedule your backups to run at regular intervals and ensure your data is always up-to-date. For instance, you can make the backups weekly, bi-weekly, or monthly. Also, be sure to test your restoration plan to ensure it works properly.

3. Use firewall

A firewall is the barrier between your network and potential internet threats. It observes and regulates inward and outward traffic based on predetermined security rules. It will block all traffic that seems harmful and allow only legitimate communications.

Using a firewall on your system will help prevent hackers, malware infections, and other cyber threats from infiltrating your system. Please either note that firewalls can be software-based, hardware-based, or a blend of both. You must also be sure to update your firewall regularly to ensure optimal protection.

4. Avoid using unsecured public Wi-Fi

Public Wi-Fi is an easy target for hackers since they lack the requisite security protection. Hackers can employ techniques like man-in-the-middle attacks to infiltrate networks on public Wi-Fi. They will try to steal sensitive user data. Therefore, while the coffee shop, school, or airport Wi-Fi can be free and inviting, it is wise to avoid it if you value your data.

To be safe, you must avoid making vital transactions or sharing sensitive data via public Wi-Fi. If you have to use the Wi-Fi, ensure you also have a virtual private network. The VPN will help encrypt your connection and ensure data safety. Alternatively, you should use mobile data connections or hotspots instead of public Wi-Fi. Taking such precautions will significantly reduce the threat posed by insecure public Wi-Fi.

5. Beware of phishing attacks

According to a Cybersecurity Insiders report, 57% of organizations experience a phishing attempt every week. The same report further notes that approx. 3.4 billion phishing emails are recorded daily. These figures show you how rampant phishing swindles have become. They tell you how easy it is to fall victim to phishing fraud.

Phishing attacks occur when attackers impersonate legitimate organizations to deceive victims into revealing sensitive information. The attacks typically happen through emails, fake websites, SMS, and calls. Phishing attackers will attach malicious links to these emails or messages. Once you click on links, it will redirect you to a phony website that is out to extort money and steal from you.

To be safe, you must learn to verify senders’ addresses and identify red flags. Look for signs such as poor grammar and urgency in the message. Most importantly, only click on the link in which you know the source.

6. Use strong passwords

Passwords are your first line of defense against attackers. Never use a weak password that makes it easy for attackers to bypass your login page and access your account. So, what makes a strong password?

While creating usernames and passwords, there are a few tips you can use to ensure you make the most secure password. Here are the features your passwords should have:

• Create a long password. A good password should be at least eight characters.
• Do not use obvious passwords. “12345” or ‘ABCDE” looks too obvious. Do not use your name, your pet’s name, or something that characterizes you.
• Use unique passwords. Blend letters, and numbers, as well as uppercase, lowercase, and special characters to make your passwords unique and hard to guess.
• Consider using password managers to generate passwords.
• Regularly update the passwords.
• Do not reuse the password. Ensure every account has its unique password.

7. Enable 2-factor authentication

Sometimes, passwords cannot be enough. There have been cases where attackers have been able to bypass passwords. 2-factor authentication can come in handy to complement passwords and boost account security. 2FA is a second layer of user verification apart from password.

2FA involves a code sent to your mobile device or email. Only you can access this code. Therefore, even when attackers manage to access your password, they will still need this code to access your system or account. The fact that they do not have the code makes things safe. With advancements in technology and the dawn of AI, 2FA has become better. Fingerprint and facial recognition systems have added a robust layer of security to the online environment.

8. Keep your software and operating system updated

Updating your software and operating system does not only help with improving performance. Software updates also help with security. They include patches to existing security vulnerabilities that attackers could easily exploit.

Making the updates is easy using the automatic update feature. The updates also help to fix bugs, introduce new features, and augment the total performance of the system. As a rule of thumb, ensure you make the updates once they are available. The longer you wait, the more vulnerable you will be.

9. Use a virtual private network where possible

A VPN can improve your online secrecy and security. The VPN encrypts the internet connection and ensures secure transmission of data. It protects internet communications against eavesdropping and interceptions.

A virtual private network is particularly important when using public Wi-Fi. It will hide the IP address and make your online activities more anonymous. For maximum security and experience, ensure you work with a reputable VPN provider that offers a strong encryption policy, no-logs policy, and reliable performance.

10. Use antimalware software

Antimalware software is among the most critical security tools you can have on your network. It will help protect your device and network against malicious software such as spyware, Trojans, ransomware, and viruses. They actively scan your system to detect, quarantine, and remove malware before it can cause damage.

In addition to malware scans, antimalware software comes with additional features. These features include web protection, email filtering, and firewall integration. These features will help you safeguard your personal information, maintain system integrity, and protect against cyber threats.

Conclusion

The best strategy to safeguard your data against cyber threats is to adopt multiple strategies. It would be more effective to have these measures work together. That way, if one fails, another will work. Remember, you can never know the strategy an attacker uses to infiltrate your system. This is why it is always good to be prepared at all times. Data breaches are real, and the best strategy is to arm yourself with multiple defenses.

When OpenAI introduced Sora, it felt groundbreaking. However, today there are many alternatives that can generate equally impressive videos.

Runway ML’s recently introduced Gen-3 Alpha follows closely on the heels of Luma AI’s Dream Machine debut. Meanwhile, Kuaishou, a Chinese TikTok rival, launched Kling, a model that competes with Sora in its impressive capabilities, capable of generating videos in 1080p high-definition resolution at 30 frames per second.

Everyone ready for another AI filmmaking tool drop later today ? The onslaught continues…

— Uncanny Harry AI (@Uncanny_Harry) June 18, 2024

Trained on both videos and images, Gen-3 Alpha will power Runway’s text-to-video, image-to-video, and text-to-image tools. It will also enhance existing control modes like Motion Brush, Advanced Camera Controls, and Director Mode, and support upcoming tools for more precise control over structure, style, and motion.

Just as OpenAI engaged with Hollywood, Runway has partnered with top entertainment and media to develop customised versions of Gen-3 Alpha. This customisation of Gen-3 models allows for greater control over style and character consistency, addressing specific artistic and narrative requirements.

Unlike OpenAI, Runway has announced that Gen-3 Alpha will soon be available to everyone.

“Gen-3 Alpha catching up to Sora in 4 months means that Omni might also be caught up by September. Anthropic might even release their agentic AI before GPT-5 is released. OpenAI should either speed up its strategy or accept falling to second place,” wrote a user on X.

Sora is Way Behind

One of the major drawbacks of Sora is that it is currently unable to generate photorealistic humans with consistent expressions and characters. However, Gen-3 Alpha excels at generating expressive human characters with a wide range of actions, gestures, and emotions.

For instance, in the short film ‘Balloon Head’ produced by Shy Heads using Sora, they replaced the human face with a balloon to overcome Sora’s limitations. Walter Woodman, the film’s creator, said that they edited the film using Premiere and After Effects.

Notably, Adobe has recently announced that it is exploring partnerships with leading AI providers such as OpenAI’s Sora, RunwayML, and Pika.

9. Prompt: A close up portrait of a woman lit by the side, the camera pulls back. pic.twitter.com/KHWvXCqUVT

— Min Choi (@minchoi) June 17, 2024

Moreover, Gen-3 Alpha is designed from the ground up for creative applications, which enables it to understand and generate a wide range of styles and artistic instructions.

Apart from Gen-3 Alpha, Luma AI claims that Dream Machine, unlike Sora, can understand the physics of the world and how things operate in the real world. Another interesting aspect of it is that it can extend existing images into videos.

“Wow. The new model from @LumaLabsAI extending images into videos is really something else. I understood intuitively that this would become possible very soon, but it’s still something else to see it and think through future iterations of,” said former OpenAI researcher Andrej Karpathy.

It's only been 4 days since Luma AI launched Dream Machine.
Memes are now becoming videos.
Here are 10 hilarious examples:
1. Disaster Girlpic.twitter.com/UXpNcB39M6

— Angry Tom (@AngryTomtweets) June 16, 2024

One of the other key differentiators is the photorealistic quality of its videos. The AI algorithms employed by Luma meticulously analyse and enhance every detail, from texture to lighting, ensuring that the final output looks almost indistinguishable from real-world footage. There are some limitations to Dream Machine, such as morphing, inserting text in videos, and camera movements.

On the other hand, the Chinese Kling model can generate videos up to two-minutes long at 1080p resolution and 30 frames per second. Known for its photorealistic output and accurate simulation of real-world physics, Kling excels in 3D face and body reconstruction, making the generated content more lifelike and expressive.

OpenAI has a GPU Advantage

A prime contributor to Luma AI’s success is AWS, which provided the company with the much-needed GPUs.

“Great to see how AWS H100 training infrastructure helped the Luma AI team reduce time to train foundation models and support the launch of Dream Machine,” said Swami Sivasubramanian, vice president for data and machine learning services, AWS.

However, shortly after becoming publicly available, the website couldn’t handle the overwhelming demand. This is where OpenAI has an edge. As the developer of ChatGPT, OpenAI has access to Microsoft Azure, providing access to the latest NVIDIA GPUs. At Microsoft Build, CEO Satya Nadella announced that they would be among the cloud providers with access to NVIDIA’s state-of-the-art Blackwell GPUs.

OpenAI recently partnered with Oracle as well to gain access to more compute.

Will OpenAI Ever Ship?

There is another interesting story associated with Luma AI. At Google I/O, Google introduced its video generation model, Veo. However, due to security concerns, Google hasn’t released the model yet.

Dan Kondratyuk, now a research scientist at Luma AI who previously worked at Google, said that he left Google because the company wasn’t shipping any products.

“I left Google to join Luma. I was part of the team that developed Veo early on but knew it would never be shipped to the masses for quite a long time, just like Sora. Not until a company like Luma forces their hand, that is (at least I hope, gimme access),” he posted on X.

Meanwhile, OpenAI has adopted a strategy of announcing products to steal the limelight from Google without actually delivering them. For instance, when Google introduced Gemini 1.5, OpenAI announced Sora on the same day. The day before Google I/O 2024, OpenAI announced GPT-4o. However, its voice feature remains unavailable.

OpenAI in a nutshell. pic.twitter.com/vmXja13OQc

— AshutoshShrivastava (@ai_for_success) June 18, 2024

On the other hand, Google, much like OpenAI, has entered the race of announcing products.

Recently, the company introduced its video-to-audio (V2A) model, which can generate audio for any video. Interestingly, this model can be paired with Veo to create shots with a dramatic score, realistic sound effects, or dialogue that matches the characters and tone of a video.

OpenAI should release Sora before it’s too late. The competition is not going to diminish.

Hollywood actor Ashton Kutcher recently praised OpenAI’s Sora, saying that creators will be able to render a whole movie using it. “I have a beta version of it, and it’s pretty amazing,” he said.

In a recent interview with The Wall Street Journal, CTO Mira Murati said that OpenAI is most likely to make Sora publicly accessible later this year.

Source: Canva

Data science has been in demand for quite some time now. Fortunately, the democratization of education has made it fairly easy to build a roadmap to learn essential technical skills.

Typically, the learning path includes building foundations comprising linear algebra, mathematics, probability, statistics, etc. along with a good grasp of at least one programming language like Python.

The Technicals

Equipped with these fundamentals, the learners become comfortable with machine learning fundamentals, understanding key algorithms – decision trees, random forests, ensembles, and time series, and eventually grasp complex deep learning algorithms.

During this journey, you will also need a good handle on concepts involving bias-variance trade-offs, the power of generalization, assumptions of algorithms, and much more. This list by no means is complete (or, will ever be), as the data science field involves continuous learning – that mostly happens through practical hands-on applications, or from learning how industry experts are doing it.

In such cases, platforms like Kaggle provide a good playground for understanding the complex nuances of building a high-performing model. Additionally, exposure to winning solutions on Kaggle not only increases their knowledge base but also enables learners to build the mindset of developing their robust models.

Beyond Tech Skills

So far, so good. But, have you noticed one thing?

The skills and the path I outlined hold no secret; they are largely available in the public domain. Everyone is learning the same approach to building skills to land their dream role in the data science domain.

This is when the reality check is necessary.

It is not just about the available AI talent but also the demand for such skills in the market. AI advancements are happening rapidly, especially since the onset of the Generative-AI era, which has prompted many organizations to reduce their workforce. Even Nvidia's CEO, Jensen Huang shared his views on future workforce and skills by highlighting that “AI will take over coding, making learning optional. AI is set to make coding accessible for everyone, reshaping how we learn to program”

Source: Immigration & Jobs Talk Show YT channel

What You Can Do?

The shifting industry landscape underscores one truth – changing times call for changing measures.

Given that the industry is witnessing a change in skill expectations, here is what you should focus on to build a stellar data science career:

• Hone the often-overlooked skill of decision-making, essential for making the trade-offs in building scalable machine learning systems.
• Build the ability to make informed decisions even in the absence of complete information, demonstrating quick thinking and adaptability.
• Building ML models requires extensive stakeholder management, implying potential friction. Master the art of stakeholder management to navigate potential conflicts and drive decisions with a compelling rationale.

Source: Canva

• Working with cross-functional teams also means that your audience might come from varied backgrounds, so building tailored communication is a big bonus.
• Most AI projects fail at the proof of concept (PoC) stage and do not even make it to production, while the ones in production struggle to show results. In short, organizations are waiting to see the returns on their AI investments. So, become that go-to person for getting things done and demonstrating the results while making progress.
• Ensure the alignment of business problems with statistical ML solutions to lead the given AI project to success. If this step goes wrong, anything downstream will not be useful.
• Innovation is a must – not just for enterprises but for all of us. Think outside the box and design innovative solutions. It is a sure-shot way to build your reputation as an expert data scientist.

The Soft Skills

Figuring out things on the fly is an art, seldom taught in classrooms. Yet, the pivotal question remains – how does one learn such skills?

There is no singular path to mastery, but here are a few starting points to develop that lens:

• Do not fear failure, instead treat challenges as opportunities to learn new things. Think of every problem statement as a gateway to learning something new in AI. It is similar to studying in university, albeit the one where you are paid for learning to make innovations come to life, instead of paying fees. Data science involves “science”, which is experimentative and involves multiple iterations to give meaningful results (and sometimes no success at all, just the learnings). These learnings accumulate over time and help you build a knowledge bank, which becomes your differentiator as you gain experience.
• Overcoming fear also means asking questions. For example, always “Start with Why?” Why are we building this? Why would our customers/stakeholders care? Why now?
• Once the "Why" behind the problem statement is clear, the "what" and "how" will follow naturally, simplifying the process of creating exceptional AI products.
• In short, in this new world where “building AI products has come down to just invoking APIs”, choosing the right problems or for that matter, inventing the right problem can pave the way for a profoundly rewarding career trajectory.

Source: builder.io

Master these skills to stand out during the interview process and build remarkable ML products that the world awaits.

Vidhi Chugh is an AI strategist and a digital transformation leader working at the intersection of product, sciences, and engineering to build scalable machine learning systems. She is an award-winning innovation leader, an author, and an international speaker. She is on a mission to democratize machine learning and break the jargon for everyone to be a part of this transformation.

More On This Topic

• Breaking Down Quantum Computing: Implications for Data Science and AI
• Data Science Minimum: 10 Essential Skills You Need to Know to Start…
• Breaking the Data Barrier: How Zero-Shot, One-Shot, and Few-Shot…
• KDnuggets™ News 22:n06, Feb 9: Data Science Programming…
• Boost Your Data Science Skills: The Essential SQL Certifications You Need
• Breaking Down AutoGPT