AI in Logistics: Emerging Startups, Challenges and Use Cases [UPDATED 2024]

In the ecosystem of logistics and supply chain management, the integration of AI promises a major shift in operations, poised to reshape the industry: the input - by changing the way raw material are acquired and transported to manufacturers, and the distribution - by transforming how products get from manufacturer to the end consumer via changing distribution strategies, customer engagement, and after-sales services.

In this report, we explore how AI has become the cornerstone of agile, data-driven logistics ecosystems, reducing costs and environmental impact, and boosting customer satisfaction. This is an updated version of our previous report, that continues to explore the uses of AI in logistics, emerging players and models, recent VC deals and exits as well as a path forward for the industry.

The key drivers behind rapid AI adoption include:

• Rising requirements for flexibility (specific pick-up points and time slots), quicker deliveries (hours not days), request for inclusivity, standardization, and sustainability.
• The requirement for transparency and end-to-end visibility across the supply chain.
• Increased availability of more structured data, new AI frontiers (GenAI), and advanced data analytics technologies.
• Expectations of future market volatility prompt diversification and continuity strategies that allow to react and adapt to supply chain shocks.
• Finally, there is always the question of pricing powered by general expectations of lower costs that put pressure on company margins and push them toward labor automation.

While there is certainly a lot of hype around AI applications in logistics and supply chain, there is also an ongoing debate about the extent of such applications as well as potential drawbacks such as mass unemployment, privacy, and integrity concerns. Before turning to risks and challenges, let’s first look at how exactly companies leverage AI to enhance their process or create new applications altogether.

Main AI Trends and Applications in Logistics & Supply Chain Management

Traditional tech and new AI techniques in computer vision, machine and deep learning, and natural language processing continue to open new pathways to effectively manage complex decision-making and operations. The AI innovation in logistics can be grouped into the following cohorts:

1. Automation of processes and tasks:
   1. Back-office automation: automating repeatable tasks like form completion, invoicing, and inventory check-ups.
   2. Brokerage automation: accelerating complex brokerage process susceptible to human error and requiring specific regulations knowledge.
   3. Data collection and storage: keeping client records up-to-date.
      
      
2. Prediction and forecasting for processes optimization:
   1. Dynamic routing: adjusting delivery routes based on real-time data like traffic, weather, as well as geospatial data to improve on-time delivery.
   2. Multi-Modal optimization: determining the most efficient mix of transportation modes (such as trucks, trains, ships, and drones) to optimize complex logistics networks.
   3. Predictive analytics: forecasting demand, inventory levels, fuel usage.
   4. Equipment maintenance: forecasting wear and tear of machinery.
   5. Demand forecasting: leveraging historical and real-time data to predict demand for route planning and resource allocation.
   6. Sustainability initiatives: reducing carbon footprint by optimizing routes.
   7. Dynamic pricing: setting the optimal price based to increase the company’s bottom line.
      
      
3. “AI-Driven Computer Vision” - logistics companies benefit from tech advancements in depth perception, 3D reconstruction, and enhancement of obscure and blurry images:
   1. Health and Safety: identifying hazards in warehouses, depots, and other logistics facilities, thus preventing accidents.
   2. Operations: detecting bottlenecks and inefficiencies via heat maps for identifying workflow patterns, unauthorized entry or intrusion.
   3. Asset management: monitoring and alerting maintenance teams, identifying asset flaws, and anomalies, and simplify defect detection.
   4. Loading and shipment processing: automating and simplifying the dimensioning process in shipping, calculating storage capacities, load planning, transportation logistics, and shipment billing, reducing the number of trips required.
      
      
4. Autonomous Vehicles:
   1. Effective operations: self-driving trucks with better efficiency, cost, safety, and reliability.
   2. Last-mile delivery: autonomous drones and ground vehicles for efficient last-mile delivery.
      
      
5. Generative AI (GenAI) - creates fresh content like text, audio, video, and more using large datasets for training where output quality depends on both the training data and user-provided prompts:
   1. Supplier selection and procurement: GenAI can provide a detailed supplier assessment and create an optimal supply list based on cost-effectiveness, product quality, reliability, and sustainability.
   2. Supply chain risk mitigation: GenAI can aid in scenario-based risk assessment by creating disruption models, including situations like supplier insolvency, wars, strikes, natural disasters, and pandemics. This ongoing analysis enables companies to develop resilient strategies for business continuity.
   3. Procurement and inventory accuracy: beyond demand prediction, GenAI can create flexible inventory policies, optimizing costs by suggesting just-in-time inventory management strategies, cutting storage expenses, and improving cash flow.
   4. Logistics coordination: GenAI can formulate contingency plans for disruptions like traffic congestion or extreme weather conditions. When provided with specific constraints, GenAI can design optimal storage and picking layouts for warehouses.
   5. Customer support: chatbots and Intelligent Virtual Assistants can provide order statuses, and estimated delivery times, and address inquiries or complaints by voice or text. Chatbots can also facilitate valuable analytics metrics, enabling the company to better understand customer needs and enhance the customer experience.
      
      

This is not an exhaustive list and as AI adoption continues and new models and tools emerge, we will see new applications as well as more granular approaches to existing ones.

Incumbent players’ AI adoption

How established players in logistics leverage AI.

While there is a lot of innovation coming from startups, established players strive to ride the wave of AI and pioneer new approaches and solutions. Often advantaged by having vast pools of data and technical infrastructure and talent necessary to adopt AI, large enterprises like DHL and Hitachi are constantly in the headlines when it comes to adopting new technologies.

For example, at DHL, AI-powered sorting robots are becoming a game-changer, increasing sorting capacity by some 40% or more. DHL has partnered with Dorabot, an AI-powered robotic solution provider capable of sorting over 1,000 small parcels per hour with 99% accuracy. DHLBots reduce missorting, removing the need for secondary sorting. Their AI forecasting and prediction models, on the other hand, help DHL to know with 90%-95% certainty that specific volumes of shipments are going to arrive in a certain facility for this day.

FedEx has invested in the artificial intelligence platform Vue.ai to make smarter supply chains and become the single unifying AI platform for large enterprises to “build entire vertically integrated stacks that are relevant to their industry.” Earlier, FedEx and Microsoft partnered to develop AI tools for enhancing logistics aimed at improving visibility and decision-making in supply chain operations, using AI-driven predictive analytics and data insights.

Hitachi is using AI in a fusion with the Metaverse to enhance railway safety. By combining AI technologies with virtual environments, they create a realistic simulation of railway operations. This allows for effective training and testing of safety procedures and scenarios, helping railway operators and staff to improve their response to critical situations.

This summer, California regulators granted permissionn for Waymo and Cruise, autonomous vehicle companies affiliated with Google, to provide around-the-clock paid taxi services in San Francisco. These two companies have the largest presence in the state, with hundreds of vehicles operating in the city. Other notable players include Apple’s secretive self-driving Project Titan with approximately 50 cars on public roads in 2022 and Amazon's Zoox which has a fleet of around 100 vehicles. Uber customers in Metro Phoenix may now be matched with a Waymo vehicle, allowing riders to experience a fully autonomous ride, with no human driver behind the wheel.

At the same time, as China advances its Level 3 automation, where a vehicle can operate independently under specific conditions, Baidu's Apollo Go is at the forefront. In June, the company secured a license to operate its autonomous ride-hailing service commercially in selected areas of Shenzhen.

AI-Enabled Startups in the Global Logistics Industry

Innovation landscape of venture-funded startups in AI in Logistics.

According to Traxn, the funding in AI logistics companies in the past 2 years reached $5.1B with top geographies including the US, India, China, Germany, and the UK. We can confirm that despite the challenging investment market in the past 18 months, many early-stage companies initially featured in our 2020 report raised Series A and later fundraising rounds.

Every logistics tech startup carries a unique approach to AI, and some of them have already operationalized several use cases. Based on public information gathered from Crunchbase, Pitchbook, and other open sources, we have compiled a list of startups that bring AI to the global logistics and supply chain industry that constitute today's Innovation Landscape in AI in Logistics 2024.

Let’s look at each category one by one.

In Analytics & Software, Portcast provides real-time updates on ocean containers to enable shippers to predict cargo arrival times accurately and set the entire supply chain in motion, cutting down on port fees, and eliminating manual work and needless delays. Portcast accomplishes this using advanced data sets and proprietary machine-learning algorithms.

Shippeo provides real-time multimodal transportation visibility, helping major shippers and logistics service providers operate more resilient, sustainable, and customer-centric supply chains. Their Multimodal Visibility Network integrates with more than 1000 TMS, telematics, and ELD systems, enabling Shippeo’s platform to provide instant access to real-time shipment tracking across all transport modes. Their machine learning algorithm allows supply chain companies to quickly anticipate problems, proactively alert customers, efficiently manage exceptions with collaborative workflows, and accurately measure CO2 and GHG emissions from supply chain transport.

Transmetrics' AI-powered platform leverages advanced machine learning algorithms and data analytics to redefine supply chain planning. It integrates seamlessly with TMS, ERP, and asset management systems, resolving data quality issues and providing access to detailed historical reporting on operational performance. By leveraging historical data combined with relevant external factors, Transmetrics builds highly accurate and reliable forecasts that can help solve challenges related to capacity, volatility, and margins by leveraging predictive planning and becoming truly data-driven.

With many fleets short on drivers and technical staff, vehicle failures are often not addressable in a timely and cost-effective manner. Uptake allows companies to discover individual vehicle risk scores and assess global fleet health, flagging potential failures and informing replacement part inventory which helps to cut maintenance costs and increase driver mileage.

In Autonomous Driving, companies like tuSimple, Gatik, and Kodiak developed autonomous trucks powered by AI/ML algorithms to make freight transportation safer, more efficient, and environmentally friendly. Kodiak, for example, is now being used to revolutionize ground autonomous military operations. Aside from its off-road driving capabilities, the Kodiak Driver drives, trusting its eyes, not its memory, just like human drive, using perceivable cues to place itself without relying solely on maps. This means that a Kodiak Driver-equipped military vehicle can drive based on satellite images alone without requiring a service member to drive a dangerous route to collect map data.

AutoX' self-driving car is capable of handling the most challenging and dynamic traffic scenarios in urban cities around the world. AutoX obtained the world's second driverless RoboTaxi permit from California and was the first company in Shenzhen and Shanghai to operate a fully driverless RoboTaxi service on public roads without any safety driver, covering the world's largest driverless area.

In Fleet Management, for efficient last-mile delivery, Wise Systems developed Route Planner, a powerful, web-based application that calculates optimal routes, factoring in the unique variables and constraints. It enables dispatchers to optimize the delivery plan for static, dynamic, or hybrid routing. The dispatcher also can customize route zoning, on-demand orders, and time windows in real-time.

Using Freightos, the world’s largest online freight marketplace, small and midsized importers can seamlessly compare, book, and manage shipments, making freight as easy as booking tickets online. The Freightos Marketplace helps importers and exporters reduce logistics spending and save time with instant comparison, booking, and management of air, ocean, and land shipments from top logistics providers.

To improve parking, and rider behavior and enable better policies, Populus empowers cities with data from shared fleets of bikes, scooters, mopeds, and cars by obtaining insights on parked vehicles and trip patterns. They also developed a Curb Manager solution for cities to measure demand, digitize parking regulations, and actively manage their curbs.

In Robotics & Warehouse Management, Southie Autonomy provides contract packagers, manufacturers, and retailers with fast and flexible robot arm automation for easy packaging assembly, kitting, and palletization. Their no-code, no-programming software platform makes it easy for any worker to deploy and changeover robot arms in less than 10 minutes.

Starship Technologies builds fleets of self-driving delivery robots designed to deliver goods locally within 30 minutes. The robots drive autonomously 99% of the time to make safe and environmentally friendly curb-side deliveries from a local hub. Delivering items such as groceries, food takeaways, and packages, Starship now operates in over 60 locations across the world and has made over 5 million autonomous deliveries to date.

To provide real-time visibility of inventory and warehouse operations, Dexory offers the only system that combines stock-scanning robots with powerful warehouse analytics. It can scan up to 15,000 locations an hour with 99.9% inventory accuracy to achieve optimal occupancy and reduce shrinkage & consistently hit SLAs.

Nimble designed a robotic system to provide automated end-to-end piece-picking, sorting, and packing solutions for e-commerce orders allowing companies to optimize warehouse locations, inventory placement, transportation strategies, and end-to-end logistics. Nimble's AI technology has picked, packed, and handled millions of items across apparel, health and beauty, footwear, electronics, consumer packaged goods, general merchandise, pharmaceuticals, and more.

In Supply Chain Management, companies like Bringg and Fareye focus on solving last-mile delivery challenges to reduce costs simplify operations, and drive brand loyalty.

When damage inspections are non-systematic and manual, they result in millions of dollars in claims each year, harming the relationship between ports, car importers, and dealers. To solve that problem, Spinframe develops vehicle inspection systems based on AI, computer vision, and machine-learning technologies that detect anomalies from the assembly line throughout the vehicle journey to the dealership and end customer. Through the Spinframe platform, you can view the fleet's condition in real-time and see a timeline of the vehicles' physical condition history.

Another player worth mentioning is a Canadian accelerator and innovation cluster Scale AI, which helps supply management companies enhance their AI-based product and service offerings and assists with AI go-to-market strategies. They have invested over $500M by reimbursing up to 50% of expenses for approved projects that target the adoption or commercialization of AI for the supply chain. All projects must be executed in Canada.

In Warehouse Automation, Hai Robotics aims to empower every warehouse and factory with logistics robots. The HaiPick solution is the world's first autonomous case-handling robot (ACR) system. It intelligently identifies desired totes and cartons, rather than the entire rack, and autonomously brings them to the workstation. HaiPick can increase warehouse operation efficiency by three to four times and storage density by 80% to 400% for clients globally.

Locus Robotics manufactures autonomous, mobile robots to support e-commerce operations for retailers and warehouse logistics providers. Its technology works collaboratively alongside workers to dramatically improve order productivity and increase fulfillment speed and throughput 2X-3X with near-perfect order accuracy. Locus eliminates unproductive walking time to significantly improve productivity and lower cycle times. Locus also improves workplace ergonomics and health safety and also helps with worker recruiting and retention.

In the table below, we assembled a list of startups that recently raised capital and are players ‘to watch’ poised for success.

M&A Activity in Logistics and Supply Chain

While the investment and M&A climate has been slow for the past 18 months, there were plenty exists in the industry.

Many M&A activities in logistics and supply chain management are driven by the need for (1) digital transformation to enhance their technological capabilities in the rapidly evolving industry (2) integration of data analytics and AI to gain insights into supply chain operations, optimize routes, and enhance overall decision-making processes (3) integration of IoT devices to improve real-time tracking, monitor the condition of goods during transit, and enhance visibility across the entire supply chain (4) expansion of global reach and optimization of logistics networks to strengthen presence in key markets (5) driving sustainability as more companies focus on integrating environmentally friendly practices and technologies to meet regulatory requirements and consumer expectations.

In the table below, we summarize most recent tech acquisitions in logistics and supply chain management.

Remaining Challenges & Path Forward

Despite massive innovation of AI in Logistics, challenges remain, some caused by limitations of the current state of AI itself. Let’s look at the major ones:

• Training, integration, and adoption of new modern technologies
  Integrating AI technology into current systems and procedures remains a future to-do task for most businesses. Many businesses are still utilizing legacy systems that may or may not be compatible with modern technology and hence will need capital and time investments to bring them up to date. While Gartner anticipates that by 2024, 50% of supply chain organizations will invest in AI applications with sophisticated analytical capabilities, organizations still face the major challenge of training and operationalizing AI systems within their business processes.
  
  The training and tuning processes demand the caliber of tech talent which might not be accessible to SMBs. In addition, the lack of robust data and proper data management processes may further put SMBs at a disadvantage impeding their ability to fully leverage AI applications.
  
  The stakes are even higher for those not adopting AI systems as they risk falling behind their tech-savvy competitors, making it challenging to reduce costs and meet customer fulfillment expectations.

• Ethics considerations of AI algorithms
  GenAI limitations, like "hallucination", biased algorithms, and misinformation, pose ethical and reputation risks in supply chain planning. Incorrect AI outputs can result in economic repercussions, systematic discrimination, and potential lawsuits. Flawed forecasting may lead to significant economic losses. To mitigate these risks, companies must promote cross-departmental collaboration to establish policies, procedures, and controls ensuring inclusive data and quality fine-tuning of AI systems.

• Privacy and cybersecurity
  Logistics cybersecurity is vital, encompassing issues like hacks, data management, GDPR compliance, and privacy laws. Integrating computer vision raises concerns about constant monitoring and potential exposure of video and image data. Organizations embracing AI in logistics need an open and proactive approach to policy and process implementation. Additionally, the connection between supply chain and national security means adversaries may target large players to infiltrate or disrupt operations, causing significant slowdowns or paralysis.

• Job replacements
  AI is poised to automate manual and repetitive tasks, particularly in logistics where roles like drivers, deliverymen, and warehouse workers are vulnerable. Research indicates varying degrees of impact and timelines for such replacements. Despite job displacement, AI is expected to create new opportunities centered around facilitating, directing, and enhancing AI capabilities. The overall effect on the job market is still uncertain, but it is evident that both employees and employers must adapt and retrain. Emphasizing areas where human intelligence excels, such as strategic thinking, problem-solving, persuasion, negotiation, and interpersonal relations, will be crucial in this evolving landscape.

• Geopolitical shocks and black swans
  Finally, the industry faces geopolitical risks and global challenges, with AI helping to mitigate these issues post-fact. Recent events, such as the war in Ukraine, have significantly impacted global food and drink industries, causing supply chain disruptions, shortages of key raw materials, and contributing to rising food prices and inflation.
  
  While AI can optimize routes, forecast inventory, and aid supplier selection based on various parameters, human intervention remains crucial for navigating the complexity of major shocks and crises. For instance, manufacturers responded to war-induced disruptions by prioritizing robust logistics management, including a shift to local sourcing for resilience and cost savings. Decisions such as banning Russian products and relying on neighboring countries for gas were driven by human considerations and support for Ukraine.
  
  Black swan events are unique and not uncommon, and AI applications currently fall short in providing comprehensive strategies for dealing with major crises. AI's limited understanding of supply chain nuances, stakeholder relationships, cultural dynamics, and the political landscape, often developed through years of industry experience, keeps humans involved in addressing complex challenges. Despite these limitations, the potential for AI to learn and adapt rapidly remains a point of interest.

In conclusion, the future of AI in logistics and supply chain management holds immense promise for efficiency, optimization, and innovation. While AI will transform various aspects of the industry, it is crucial for businesses to balance automation with human expertise. Collaboration, adaptability, and a strategic approach to integrating AI technologies will be key to navigating the evolving landscape, ensuring resilience, and unlocking the full potential of this transformative technology.

DRL Cases in Logistics & Smart Mobility

We at DataRoot Labs have helped several startups and enterprises build their AI products. Below you can read about how we used the power of AI to deliver AI-powered MVPs to logistics and smart mobility-related startups.