Get 4 new trends every week
Delivered to your inbox each Tuesday
See trends before they become trends
Trusted by 65,000 people
Insights, expert analysis, and more
Loved by over 65,000 people at companies like:

2021's Biggest Technology Trends

This is a list of important technology trends to keep an eye on in 2021.

In this report we’ll cover key trends in the tech space, including:

  • New use cases for game engines
  • Industrial applications for VR
  • Delivery and logistics tech
  • Lots more

We'll cover each of these trends in depth. And give you a peek into what may be coming next.

1. Game Engines Are Everywhere

Global spending on video games is expected to hit $159 billion this year, and exceed $200 billion by 2023.

global game market forecast-min.png
The global video game market is set to see continued growth through 2023.

Two of the most popular game engines that power those games are Unity and Unreal Engine.

Game engines provide a foundational layer of graphics, physics and other elements. That way, developers don’t have to create everything from scratch.

Unity Technologies alone made $541.8 million last year chiefly from their game engine. (The company went public in 2019 and has a market cap of $37.5B as of this writing.)

In fact, the Unity game engine was used by 53% of 2019’s top 1,000 games in the iOS App Store and Google Play Store.

For context, 48% of video game revenue comes from mobile gaming, which is also the fastest-growing market segment.

But Unity isn’t just for mobile games: it’s used by 94% of the largest game studios by revenue.

While Unreal Engine, owned by privately-held studio Epic Games, has been used internally for giant hits like Fortnite and Gears of War.

Searches for "Epic Games" over 5 year-span.

As well as by other game studios for thousands of titles including the Tony Hawk’s Pro Skater series, Valorant, and the Borderlands franchise.

Unreal Engine 5 was just announced a few months ago, but several games are already being developed with it for the next-gen Playstation 5 and Xbox Series X and S consoles.

To use Unreal Engine, game developers must pay a 5% royalty on worldwide gross revenue above $1 million.

While Unity doesn’t require royalties, but instead charges monthly fees ranging from $40/month to $150/month per user.

But game engines aren’t just for games.

In fact, so many things run on game engines that the makers aren’t even calling them game engines anymore. Unity uses the term “real-time development platform” while Epic describes the Unreal Engine as a “real-time 3D creation platform”.

Another way these platforms are used is for creating “digital twins”.

Digital twins are digital likenesses of physical objects or locations. They can be redesigned, tweaked or manipulated much more quickly and affordably than the real thing.

Searches for “digital twin” have increased 253% over the last 5 years.

For example, the Hong Kong International Airport is using a Unity-powered digital twin to help build an extension.

While the digital twins of Helsinki, Tampa and Shanghai have all been created with Unreal Engine for various purposes.

Here are some other non-gaming ways these two development platforms are being used:

Toyota uses Unity in numerous ways, from digital engineering to conducting VR training, creating incredibly realistic renders for its Lexus car configurators and more. (Unity is also used by BMW and Volvo.)

Disney’s The Mandalorian is produced in part with Unreal Engine, and so were Rogue One and Finding Dory. While The Lion King, The Jungle Book, Ready Player One and Blade Runner 2049 all used Unity.

The Weather Channel uses Unreal Engine for its mixed reality video segments.

Wayfair’s AR app uses Unity to accurately render products in 3D.

Lockheed Martin uses Unity for virtual prototyping and testing, saving millions of dollars in the process.

NASA has used Unreal Engine to help train astronauts, NATO uses it for military simulations and training, and Precision OS uses it to train orthopedic surgeons.

Unity and Unreal Engine aren’t the only options around, either.

Cocos is another popular development platform, with a focus on less-intensive 2D games. According to the company, the Chinese game engine was at least partially behind 5 of the top 10 games in China as well as 30% of mobile games worldwide in 2018.

Another is Godot, a not-for-profit, free and open source game engine that has launched several related training courses via Kickstarter.

Search interest in the Godot engine has been climbing quickly.

While Sony has PhyreEngine, which has been used for hundreds of titles (especially Playstation ones).

Other smaller competitors include Construct, GameMaker and Twine, which power an estimated 12.3%, 11.0% and 6.2% respectively of hobbyist and indie PC games.

As well as Source, CryEngine and Gamebryo, which are behind an estimated 4%, 3.5% and 3.2% of AAA PC games.

And some of the biggest game studios have their own game engines for internal use. Like Frostbite by Electronic Arts, id Software’s id Tech, Bethesda’s Creation Engine and Rockstar’s RAGE.

Related to the game engine market, there’s also a sub-industry of “middleware”.

Middleware is software that can be tacked onto a game engine to improve its functionality in some specific way.

Melbourne’s FMOD, the UK’s Fabric, and the Canadian Wwise are all forms of middleware that focus solely on audio.

Other types of middleware include Havok AI for AI logic, Umbra for rendering optimization, and SpeedTree for modeling trees and plants.

2. Niche AI Services

OpenAI’s breakthrough language model GPT-3 made a “shock and awe” level of impression this year.

Search growth for OpenAI since 2016.

Since users started getting access in June 2020, it’s been used to write code, create blog posts that seem written by humans, and even write fiction and poetry.

tweet about GPT-3-min.png
GPT-3 can code based on natural language input.

To many, this looks like the future of writing, programming, customer service and more.

But if the future will be driven by AI, then the future is already here.

It’s estimated that in 2018, artificial intelligence created global business value of $1.2 trillion.

Gartner forecasts that in 2021, that number will rise to $2.9 trillion contributing 6.2 billion hours of worker-equivalent productivity.

Meanwhile, revenue from AI software sales is expected to grow from $10.1 billion in 2018 to $126 billion in 2025.

Those figures aren’t based on the holy grail of “artificial general intelligence”, which many experts think is still decades away (and some believe will never happen).

Instead, they’re based on purpose-built AI being applied with a more narrow focus.

Niche-focused AI businesses have been growing behind the scenes for years, helping businesses, schools and governments get things done more effectively and efficiently.

For example:

Tractable helps car insurance agencies settle claims faster by using machine learning to assess damage and estimate repair costs.

The London-based startup has raised $55 million since being founded in 2014. website-min.png
Shot from Tractable's website.

EnsoData assists clinicians at sleep centers in assessing patients’ sleep disorders and providing treatments.

After raising a $9 million Series A round in June, the company plans to broaden its scope to other health issues as well.

Zest AI offers AI-driven portfolio management, credit decisioning and more to banks and financial service providers.

The company quotes the CEO of Discover as saying “Banks that fail to invest in machine learning will end up fundamentally uncompetitive in a couple of years.” Zest AI has raised $217M in funding since it was founded in 2009.

Clearbit aggregates and analyzes customer information. Then brings it all together into companies’ CRMs in order to make sales or provide better customer service to reduce churn.

Clearbit was last valued at $250M in early 2019.

MonkeyLearn offers “no-code AI” that lets businesses train machine learning models for text-analysis purposes. Like understanding intent or sentiment. And it’s available with approachable, monthly SaaS-style pricing.

According to their homepage testimonials, MonkeyLearn has been used by Drift to create a new feature for their customers, and it’s been used by Clearbit (above) as an integral part of their business. MonkeyLearn raised a $2.2 million seed round in July.

Searches for MonkeyLearn over the last 5 years.

Dor sells a foot-traffic analytics service for retail storefronts and other physical spaces.

Using thermal-detection devices, they show owners how many people have been visiting, when, at which locations, and what they’ve been doing inside those locations. They raised an $8.5 million Series A round in 2018. helps brokerage firms facilitate equity trades. (And it helps stock traders improve their performance.)

The Israel-based startup has raised $10M since it was founded in 2015.

Cylance offers AI-driven cybersecurity services for customers as diverse as high schools, banks, hospitals and the Sydney Opera House.

Founded in 2012, Cylance was acquired by BlackBerry in 2019 for $1.4 billion.

Not to say that broader AI services don’t exist.

After all, not all businesses’ use cases are covered by pre-developed specialized AI.

Many of the biggest tech companies offer all-purpose AI platforms. (And their technology is often the backbone of the niche AI providers.)

Like Amazon’s SageMaker, Google’s Cloud AutoML, Microsoft’s Azure AI, IBM’s Watson Studio and Alibaba’s Machine Learning Platform for AI.

Outside of those big names, companies like, CognitiveScale and RapidMiner also offer general services to help organizations create their own AI solutions.

What kinds of organizations do they help? sells to 3M, Shell, Enel and the U.S. Airforce.

CognitiveScale’s clients include Morgan Stanley, NBCUniversal, Dell and Keller Williams.

While RapidMiner serves BMW, Cisco, HP and Transport for London.

(As well as many others.)

3. Industrial AR and VR

VR has been a hot technology trend for years now and it largely hasn't lived up to the hype.

And AR’s first massive breakout hit was the 2016 mobile game Pokémon Go. (Which has earned nearly $4 billion over its lifetime so far.)

But as interesting as these technologies are for gamers, their industrial uses are even more promising.

First though, let’s talk about the elephant in the room:

VR and AR startups haven’t lived up to their expectations so far.

A pioneer in the space, Y Combinator-backed AR headset startup Meta Company declared insolvency and sold its assets in January 2019.

Osterhout Design Group faced a similar fate, closing down a few years after raising $58 million in 2016.

And while VR startup Magic Leap is still breathing, it appears to be on life support. After attracting funding from Google, Andreessen Horowitz, Kleiner Perkins and others, it seemed poised to change the world in a variety of ways.

That hasn’t panned out yet. Not for Magic Leap, and not for any other mixed-reality technology business.

At least, not as much or as quickly as they were hoping.

But these technologies are starting to catch on. And to make a difference in multiple industries.

Microsoft’s HoloLens 2 AR headset is just one example.

Searches for Microsoft’s Hololens 2.

Toyota uses the HoloLens 2 to build cars faster, inspect paint jobs, install equipment and more.

Mattel uses it to design new toys.

And other companies use it to help repair broken machinery, view schematics and more.

The Sequoia-backed AR startup Mira is also industrially oriented.

But their Prism Pro headset is technologically simpler than the HoloLens: it consists of a helmet with goggles and a head-mounted camera, requiring the addition of a smartphone to function.

This reduces its price, increasing its rate of adoption.

Mira CEO Ben Taft claims their growth in adoption has allowed Mira to “onboard more customers to our platform that are using our device every single day than companies like Magic Leap that have raised literally hundreds of times [Mira’s] funding.”

As we all acclimate to a remote workplace, VR is also bringing teams together again.

The VR teleconferencing startup Spatial has raised $22 million over the past two years from some notable investors. And search interest in the app has more than tripled in the past year.

Spatial’s focus?

Virtual collaboration rooms that support any headset or AR glasses you already have. Or even if all you have is a webcam.

Other recent entrants to this space include:

- Sine Wave Entertainment. This London-based developer was previously focused on virtual reality gaming. But they recently pivoted to a VR social hub for remote teams called Breakroom.

- Space’s unique angle is that you can use it to add VR to Zoom, Google Meet, Skype and more.

- MeetinVR, which reportedly is being or has been tested by more than 2,000 organizations including numerous Fortune 500 companies.

- Lenovo recently announced their "ThinkReality A3",  a brand of "enterprise smart glasses".

VR and AR are also being used for industrial training.

Texas-based Interplay Learning offers VR-powered job training for skilled trades. Adult students get virtual hands-on experience in plumbing, electrical, HVAC maintenance, solar panel installation and more. Interplay Learning offers a SaaS pricing model for individuals and businesses.

While Labster provides VR laboratories to schools and employers, offering access to virtual million-dollar labs for single digits per user per month.

Interest in the Danish startup has grown rapidly this year, as you can see below.

Search interest in Labster increased rapidly in 2020.

Shrinking equipment prices are helping to drive this tech trend.

Standalone VR headsets are now cheaper than ever. And they’re expected to drop further to $200 by 2023.

While mobile-based VR housings which rely on your phone to provide the computing power and display are already far cheaper than that.

Compute power has gotten less expensive as well.

Microsoft, Amazon and Google have all contributed, of course.

But smaller competitors like GridRaster are also emerging to power VR and AR experiences.

4. Autonomous Checkouts

By 2024, an estimated 10,000 stores will have autonomous checkouts globally — up from just 350 in 2018.

autonomous checkout growth chart-min.png
Autonomous checkout adoption rates.

And smart checkout technologies are expected to process $387 billion in transactions in the year 2025. Up from $2 billion this year.

Walmart, Kroger, Wegmans and Price Chopper all offer (or are testing) app-based self-checkouts.

But mobile self-checkout requires shoppers to scan each item with their smartphone, pay through an app, and then show their digital receipt to someone before they leave the store.

Autonomous checkout, on the other hand, eliminates manual scanning and payment altogether.

Shoppers simply collect the items they want from store shelves, walk out of the store and the payment happens in the background.

(Though sometimes there are other short steps, like scanning a QR code as they enter the store.)

Autonomous checkout is attractive to retailers because it promises lower overhead: stores spend an estimated $50 billion per year on cashiers’ wages in the United States alone.

And because autonomous checkout includes real-time inventory tracking, another benefit to retailers is streamlined inventory ordering and restocking.

The payback period for autonomous checkout hardware (cameras, sensors, etc.) is currently estimated at about 2 years. But like with other technology, the cost should fall quickly over time.

And the benefit to customers?

Simple: no checkout lines. 60% of consumers say that long lines for payment and checkout are a major pain point in stores.

The lack of lines (and cashiers) is also attractive for health reasons.

According to a recent study, 58% of consumers plan to avoid crowds and 46% plan to spend less time inside stores even after the COVID-19 pandemic has passed.

The best-known leader in autonomous checkout is Amazon.

The company appears to be developing an entire shopping-and-checkout experience to sell to brick-and-mortar retailers. (Like it did with Fulfillment By Amazon for online retailers and AWS for the rest of the web.)

Amazon opened its first cashierless convenience store, Amazon Go, to the public in January 2018. The project has expanded since then to 26 locations, plus a full-size Amazon Go Grocery store which opened in Seattle in February.

Amazon Go stores use entrance gates to identify customers, shelf sensors to detect which products are selected, and hundreds of cameras to see who grabbed what. With AI powering everything in the background.

In March this year, the company started selling this “Just Walk Out” technology stack to other retailers.

(So far, the only confirmed retailer to adopt the technology is a chain of airport convenience stores. But Amazon claims it has other clients lined up. And it’s rumored that Amazon itself will be rolling out the tech to its Whole Foods stores starting next year.)

Then in July, Amazon announced the Dash Cart: a smart shopping cart that uses computer vision and sensors to identify items as they’re placed inside.

This adds partial autonomous checkout capability to stores that don’t have the advanced sensors, cameras and equipment needed for a full setup.

Amazon isn’t selling the Dash Cart to other retailers (yet), but it’s being tested in at least one Amazon Fresh grocery store.

dash cart-min.png
The Amazon Dash Cart, with built-in cameras, sensors and touchscreen.

Finally, the most recent development in Amazon’s retail plans was announced two weeks ago.

That’s when they unveiled Amazon One: a contactless identification and payment system that can be used at checkouts or entrances to stores. As well as at event locations, stadiums, offices and more.

(Amazon One is based on palm biometrics, which have also been used in South Korea for several years.)

But Amazon isn’t the only company pushing autonomous checkout forward.

Their biggest domestic competitor is Standard Cognition, a San Francisco-based startup that was valued at $535M.

The company doesn’t use shelf sensors or entrance gates, but instead installs dozens of overhead cameras to identify customers and products.

Payment happens automatically in the background if you use their companion app (or you can pay at in-store kiosks).

Standard Cognition is installing its autonomous checkout tech in multiple Circle K convenience store locations, as well as at the Worcester Red Sox minor league baseball team’s new stadium.

And earlier this year, the company acquired an Italian startup called Checkout Technologies as part of its international push.

Other autonomous checkout competitors include:

Zippin, which raised a $12M Series A late in 2019. Zippin’s cashierless checkout tech is similar to Amazon Go’s, using cameras, gates and shelf sensors.

Zippin announced a deal to provide the technology to Lojas Americanas, Brazil’s largest retailer. The first location opened to the public in Rio de Janeiro in April 2019.

Unlike Amazon, Zippin also offers a store-in-a-box product called the Zippin Cube. The 300-500 square-foot Zippin Cube is almost like a walk-in vending machine: it can be placed inside an existing structure, such as the lobby of an office building.

zippin cube.jpg
The Zippin Cube.

AiFi, which (like Zippin) offers autonomous checkout tech for retrofitting its own modular store-in-a-box called the NanoStore.

Thanks to partnerships with chains like Loop Neighborhood, Żabka, Albert Heijn and Carrefour, the company is reportedly planning to have 330 autonomous checkout locations by the end of next year.

AiFi has raised $15M to date with backing from Greylock and others.

Grabango, which recently announced that a Giant Eagle store location was opening with its cashierless tech. And the company deployed the world’s largest checkout-free grocery store.

Grabango has raised $32.2M to date.

Storelift, announced that it had launched its first two autonomous stores. The French startup uses shipping containers as the structure for its “Boxy” convenience stores.

Searches over time for French startup Storelift.

Trigo, a Tel Aviv-based startup currently piloting its cashierless checkout tech in the Israeli supermarket chain Shufersal.

Inokyo, whose tech does require a one-tap payment checkout but no manual product scanning.

Tiliter, a Sydney-based startup with product recognition tech that can be used for autonomous checkout, smart produce scales, or mobile checkout apps.

As well as smart shopping cart startups Veeve, Storewide Active Intelligence and Caper.

While 7-Eleven is testing its own autonomous checkout technology, developed in-house.

5. Delivery Robots Close the “Last Mile”

The autonomous “last mile” delivery market is currently valued at $12 billion. By 2030, it’s expected to grow to $91.5 billion.

last mine total market value-min.png
The global "last mile" delivery market.

One big reason: ecommerce retailers have been making delivery times shorter and shorter.

Since announcing Amazon Prime in 2005, Amazon has been leading this charge.

Search growth for "Amazon Prime" has been slow and steady since 2016.

The company said it was transitioning from two-day fast shipping to one-day as the new standard Prime option. And now, millions of items are available for one-day or same-day shipping. With tens of thousands available within 1-2 hours.

Wal-Mart, Target and others have followed suit, offering next-day or same-day delivery options in at least some locations.

Building new warehouses and “mini-fulfillment centers” closer to urban centers is one way to help achieve these speeds. But as Jeff Bezos has said, customers will always want fast(er) delivery.

Delivery robots look like the next big strategy for fulfilling that desire.

Including ground-based solutions like the Amazon Scout, a six-wheeled bot the company recently started testing.

The Amazon Scout delivery robot.

Earlier this month, Amazon also unveiled an all-electric delivery van (to be made by Rivian). While it will be operated by human drivers initially, the plan is to make it at least partially autonomous.

But for now, lots of other companies are working on smaller delivery robots that are more like the Amazon Scout. These robots are also known as SADRs: sidewalk autonomous delivery robots.

For example:

Starship Technologies delivers food at colleges and elsewhere. Customers can pay a $1.99 delivery fee, and in at least some cases restaurants also pay a fee of 20% of the order price. Customers can place an order through the Starship app, which is also used to identify themselves to the robot so it will unlock and open. Each robot can carry up to 20 pounds.

Starship has made over 500,000 deliveries so far. It has a fleet of dozens of robots on each college campus, including Arizona State University, The University of Texas at Dallas and Purdue University. Outside of college campuses, Starship delivers take-out meals and groceries in a number of cities including Mountain View and Modesto, California and Milton Keynes, England.

The San Francisco startup was founded by two Skype co-founders in 2014 and has raised $82.2M in funding.

Starship Technologies delivery robot-...
The Starship Technologies delivery robot.

Refraction AI makes the REV-1, a 3-wheeled vehicle that’s about the size of a bike. It has space for about 4-5 grocery bags.

One way the startup is differentiating itself is by making sure it can deliver just as well in the snow as it can in sunny weather.

Refraction AI is currently delivering food in Ann Arbor, Michigan — where it’s seen a 3x-4x increase in orders since the start of the COVID-19 pandemic. The outbreak prompted the company to add disinfecting UV lights to the interior bag compartment.

So far, Refraction AI’s monetization model is to charge customers a $3 fee or restaurants 15% of the order cost for delivery.

Nuro makes an on-road autonomous delivery vehicle about the size of a subcompact car. The startup was founded by two ex-Google engineers and raised $940M from Softbank at a $2.7B valuation.

Nuro has struck delivery agreements with Domino’s and Kroger, but according to the CEO its tech isn’t yet ready to start delivering pizzas and groceries.

Search interest in Nuro has been growing over time.

Robby Technologies partnered with PepsiCo to launch the Snackbot: a cooler-sized robot that delivers snacks on at least one college campus. Though there hasn’t been much news about the Y Combinator-backed startup this year.

AnyBotics makes dog-sized robots that can climb stairs — since they have legs.

There’s also BoxBot which is backed by Toyota, Savioke for interior locations like hospitals and hotels, Postmates Serve, TeleRetail and Eliport.

And that’s just for ground-based delivery.

Then there are aerial delivery drones.

Searches for “delivery drone” over time.

Alphabet’s Wing and UPS Flight Forward were the first two aerial drone services to be granted FAA approval for U.S. commercial deliveries in the U.S. (For beyond the operators’ line of sight.)

Several months later, the companies both made their first customer deliveries.

Wing’s first delivery was of a few products from Walgreens. While UPS’s first major customer is CVS.

Wing is currently available in Christiansburg, VA in the U.S., Helsinki, Finland, and Canberra and Logan in Australia. It has partnered with a handful of local retailers, bakeries and restaurants in each location, and also offers FedEx deliveries in select parts of Christiansburg.

In April, UPS Flight Forward announced plans to deliver prescriptions from CVS to The Villages: America’s largest retirement community, which 135k people call home.

UPS has been partnering with Matternet for its drones and logistics system so far.

But UPS also announced a new partnership this year: German startup Wingcopter is developing package-delivery drones for the shipping giant. The new drones are capable of both vertical and horizontal flight, which should allow for longer-range deliveries.

Then there’s Amazon.

Amazon reportedly shipped 415 million packages in July, delivering 66% of its own packages that month. The company has been rapidly moving its delivery fulfillment in-house and is expected to eventually surpass UPS and FedEx in U.S. package volume.

Amazon’s Prime Air drone delivery fleet was granted FAA approval in August for deliveries beyond the visual line of sight.

With Prime Air, Amazon plans to deliver packages to customers within 30 minutes of ordering. Like Wingcopter’s drones, Amazon’s are capable of both vertical take-off and landing and horizontal flight.

FedEx has its own plans for autonomous delivery (aside from its partnership with Wing for small local packages). The company is also developing unmanned air cargo planes.

To do so, FedEx revealed that it was working with startup Reliable Robotics (which raised $33.5M over the summer).

FedEx is reportedly starting with a single semi-autonomous 14-passenger airplane. That’s smaller than many of the cargo airplanes the company uses, but much larger than last-mile delivery drones.

With 35% of the world’s trade value transported by air, making air freight autonomous or semi-autonomous at scale could make a massive impact.

However, FedEx says the transition will likely take decades to scale. Compared to last-mile delivery drones, autonomous full-sized aircraft obviously present higher risks.

Other players in the aerial drone delivery space include:

Zipline, which is focused on healthcare product delivery in Rwanda and has made over 65k commercial deliveries so far.

Indian food delivery company Zomato.

Flirtey, which made the first commercial delivery to a private residence in 2016 (from 7-Eleven).

Airbus Skyways, which completed the world’s first shore-to-ship drone delivery.

Boeing’s larger “cargo drones”, which are in development but will reportedly carry up to 500 pounds.

Alibaba, which started testing drone delivery in China in 2015.

China’s, which has deployed drones to good effect during the Coronavirus pandemic.

As well as DHL, Rakuten, Flytrex, Drone Delivery Canada, Manna Aero, Uber and Skycart.

Looking Ahead

That concludes our list of influential technology trends for 2021 and beyond.

Tech is one of the most interesting spaces to be involved with (largely because it's always changing).

Hopefully this list of tech trends helped give you a glimpse into what's coming next.

Last Updated: 
January 12, 2021
Josh Howarth
Co-founder of Exploding Topics.

Get Weekly Trending Topics

Our best Exploding Topics, plus expert analysis.

548 Market St. Suite 95149
San Francisco, California
© 2021  Exploding Topics