BMW’s Alexa integration gets it right

BMW will in a few days start rolling out to many of its drivers support for Amazon’s Alexa voice assistant. The fact that BWM is doing this doesn’t come as a surprise, given that it has long talked about its plans to bring Alexa — and potentially other personal assistants like Cortana and the Google […]

BMW will in a few days start rolling out to many of its drivers support for Amazon’s Alexa voice assistant. The fact that BWM is doing this doesn’t come as a surprise, given that it has long talked about its plans to bring Alexa — and potentially other personal assistants like Cortana and the Google Assistant — to its cars. Ahead of its official launch in Germany, Austria, the U.S. and U.K. (with other countries following at a later date), I went to Munich to take a look at what using Alexa in a BMW is all about.

As Dieter May, BMW’s senior VP for digital products told me earlier this year, the company has long held that in-car digital assistants have to be more than just an “Echo Dot in a cup holder,” meaning that they have to be deeply integrated into the experience and the rest of the technology in the car. And that’s exactly what BMW has done here — and it has done it really well.

What maybe surprised me the most was that we’re not just talking about the voice interface here. BMW is working directly with the Alexa team at Amazon to also integrate visual responses from Alexa. Using the tablet-like display you find above the center console of most new BMWs, the service doesn’t just read out the answer but also shows additional facts or graphs when warranted. That means Alexa in a BMW is a lot more like using an Echo Show than a Dot (though you’re obviously not going to be able to watch any videos on it).

In the demo I saw, in a 2015 BMW X5 that was specifically rigged to run Alexa ahead of the launch, the display would activate when you ask for weather information, for example, or for queries that returned information from a Wikipedia post.

What’s cool here is that the BMW team styled these responses using the same design language that also governs the company’s other in-car products. So if you see the weather forecast from Alexa, that’ll look exactly like the weather forecast from BMW’s own Connected Drive system. The only difference is the “Alexa” name at the top-left of the screen.

All of this sounds easy, but I’m sure it took a good bit of negotiation with Amazon to build a system like this, especially because there’s an important second part to this integration that’s quite unique. The queries, which you start by pushing the usual “talk” button in the car (in newer models, the Alexa wake word feature will also work), are first sent to BMW’s servers before they go to Amazon. BMW wants to keep control over the data and ensure its users’ privacy, so it added this proxy in the middle. That means there’s a bit of an extra lag in getting responses from Amazon, but the team is working hard on reducing this, and for many of the queries we tried during my demo, it was already negligible.

As the team told me, the first thing it had to build was a way to switch that can route your queries to the right service. The car, after all, already has a built-in speech recognition service that lets you set directions in the navigation system, for example. Now, it has to recognize that the speaker said “Alexa” at the beginning of the query, then route it to the Alexa service. The team also stressed that we’re talking about a very deep integration here. “We’re not just streaming everything through your smartphone or using some plug-and-play solution,” a BMW spokesperson noted.

“You get what you’d expect from BMW, a deep integration, and to do that, we use the technology we already have in the car, especially the built-in SIM card.”

One of the advantages of Alexa’s open ecosystem is its skills. Not every skill makes sense in the context of the car, and some could be outright distracting, so the team is curating a list of skills that you’ll be able to use in the car.

It’s no secret that BMW is also working with Microsoft (and many of its cloud services run on Azure). BMW argues that Alexa and Cortana have different strengths, though, with Cortana being about productivity and a connection to Office 365, for example. It’s easy to imagine a future where you could call up both Alexa and Cortana from your car — and that’s surely why BMW built its own system for routing voice commands and why it wants to have control over this process.

BMW tells me that it’ll look at how users will use the new service and tune it accordingly. Because a lot of the functionality runs in the cloud, updates are obviously easy and the team can rapidly release new features — just like any other software company.

Alexa Cast Will Allow Users To Throw Audio To Alexa-Enabled Devices Like AirPlay And Google Cast

Amazon Alexa Cast will soon allow users to throw audio to Alexa-enabled devices like Apple AirPlay and Google Cast. Here are the details. [ Continue reading this over at RedmondPie.com ]

Amazon Alexa Cast will soon allow users to throw audio to Alexa-enabled devices like Apple AirPlay and Google Cast. Here are the details.


[ Continue reading this over at RedmondPie.com ]

Amazon’s newest Alexa Fund recipients are less consumer-focused

Amazon is today announcing the new batch of startups joining its Alexa Accelerator program, powered by Techstars. Members of last year’s Amazon’s Alexa Accelerator program, which backs companies developing new experiences using voice-based technologies, went on to raise over $10 million in venture capital following their participation in the program, says Amazon. The new group […]

Amazon is today announcing the new batch of startups joining its Alexa Accelerator program, powered by Techstars. Members of last year’s Amazon’s Alexa Accelerator program, which backs companies developing new experiences using voice-based technologies, went on to raise over $10 million in venture capital following their participation in the program, says Amazon. The new group includes startups focused on business use cases, STEM education for kids, accessible technology, and more.

The idea behind the accelerator is to help fuel early-stage companies developing for voice, giving Amazon an equity stake in the businesses.

The teams will participate in a three-month long accelerator program that culminates on October 9th with Demo Night, where they’ll present their business to venture capitalists and angel investors, and present their new Alexa experiences.

Amazon says it received hundreds of applications from 44 countries around the world for the 2018 program, and narrowed it down to nine it believes have the most potential.

During the accelerator program, the companies will improve their products, refine their business model, and develop for Alexa, while receiving mentorship from both Techstars and Amazon, as well as the broader Seattle community.

This year’s batch includes participation from the following:

Blutag

Blutag seems especially relevant to Amazon’s interests, as its company is helping stores create voice-based shopping experiences for their customers. It aims to enable retailers to build a voice-based store without coding, allowing customers to shop by asking Alexa for a particular product, then receive personalized product suggestions over text or email.

Conservation Labs

This startup is operating in the smart home space, offering a produce that connects to the home’s main water line to monitor household water use, in order to help homeowners gain money-saving insights and detect leaks.

HelixAI

HelixAI is taking Alexa to scientific laboratories. Not to be confused with genetic services marketplace Helix, this startup’s HelixAI digital assistant can respond to natural language queries to provide scientists and other others in lab settings with real-time information about their operating procedures, lab safety information, workflow and processes, and reference information. For example, you can ask HelixAI things like “what’s the boiling point of benzene?” or “What about the cut site for the restriction enzyme EcoRI?”

Imageous

Imageous is expanding Alexa’s smart home capabilities to the “smart building.” Its smart facilities AI assistant for occupants of commercial buildings brings the benefits of AI technology to building operators. The AI can take advantage of system data (environmental), social data from the occupant population (if they’re reporting they’re hot, cold or comfortable), and external data sources (e.g weather or traffic data), to optimize the building for energy use and comfort with a focus on efficiency and cost savings.

Jargon

Jargon is offering an on-demand translation service that removes language barriers by combining technology with human assistance.

Novalia

Novalia offers a Bluetooth platform connected to paper-thin self-adhesive touch sensors that capture data through touch, in order to create immersive, touch-based experiences, Its audio platform then responds to touch, and turns it into audio through a surface sound actuator or line out. The company has worked with a number of brands on digital signage, touch-based posters, and other projects.

Presence AI

This company is developing AI-powered conversations for small businesses to replace phone calls for things like bookings. Currently it operates over text message, but an Alexa integration could translate this to voice. (A less troublesome version of Google’s Duplex, perhaps, as it doesn’t try to impersonate a human.)

Unruly Studios

Boston-based Unruly is combining STEM education with physical activity by building programmable, electronic floor tiles that kids can code, then jump on, and run around on to play interactive games. The startup includes former engineers from Hasbro, iRobot, Mattel and Rethink Robotics.

Voiceitt

This company is working to make voice technology accessible, with the development of Automatic Speech Recognition technology (ASR) that allows people with severe speech impairments to communicate and be understood by voice. Customers train the software to understand their unique pronunciations, which it then translates into normalized speech output in the form of audio or text. The system can also be used to help people have face-to-face conversations.

(TechCrunch coverage: Voiceitt lets people with speech impairments use voice-controlled technology)

A number of these companies in this cohort are more focused on supporting businesses, rather than consumers, using voice technology. That’s not surprising given Amazon’s recent interest in putting Alexa in the office and in hotels, for example.

Amazon’s Alexa Fund backs the participating startups with an initial $20,000 funding in return for a six percent equity stake. The startups also have the possibility of receiving another $100,000 as a convertible note. 

However, there have been some concerns that along with the rewards, there are also risks for startups joining Amazon’s program. As The WSJ pointed out as did The Information, some entrepreneurs have taken a wary view of working with Amazon’s VC arm – especially after it led the Series A for home videoconferencing startup Nucleus, then proceeded to directly compete with it with the subsequent launch of the Echo Show.

But on the flip side, startups get an early peek at Amazon’s Alexa roadmap, and access to Amazon staff for help in developing Alexa skills.

Last November, Amazon announced an additional $100 million in venture capital for the fund targeted at international investment opportunities. Past Alexa Fund portfolio companies have included ecobee, TrackR, Rachio, Toymail, Ring (which Amazon acquired), Sphero, Vesper, Owlet, and many more.