20世纪70年代以来，伴随计算机应用的日益广泛，以及人们对现代技术认知的指数级增长，人工智能逐渐被认为是引领各行各业走向未来的重要基础。而无论是为顾客提供无人收银体验的Amazon Go，还是使用IoT技术改造重要基础设施、确保各项工具处于良好工作状态的医院等，种种事例似乎无一不在证明这点。
长期以来，人们普遍认为人工智能相关应用只适用于财力雄厚的大型零售企业，但当下也有一些提供AI数智化零售解决方案的初创公司，通过研发创新帮助大大小小各类规模的线下零售门店快速迈向未来。云拿科技便是这样一位引领者与赋能者。
近期，致力于整合前沿零售解决方案与独家商业内容、帮助零售企业实现突破性增长的创新媒体——Retail IT Insights采访了云拿科技CEO Jeff Feng，双方围绕“人工智能如何加速零售业创新”，深入探讨了人工智能在零售业中的现状、潜力以及在行业转型后的前景。以下是本次采访的精彩内容：
1、人工智能和机器学习在零售业的现状如何？相较更早期的人工智能应用，我们取得了多大的进步？
Jeff：目前，人工智能在零售业的应用已经十分广泛，可以切入顾客体验、营销、商品管理等各零售环节提供解决方案，例如通过对顾客需求及偏好的洞察，帮助经营方创建不同的、更高效的营销策略或是选品策略等。但人工智能和机器学习的潜力远不止于此，其已发展到能帮助任何规模的传统门店转变为高度数字化、智能化、无需人工收银的智慧门店，这也将为整个零售业带来极大的潜力。
Jeff: Right now, AI in retail has numerous applications and can be utilized to provide solutions for customer experience and service, marketing and nearly every function within the retail space. Most of these functions are about providing insights to create different retail strategies - marketing campaigns and similar functions - that are more effective, but the potential for AI and machine learning goes much further than just insights. It’s progressed to the point now where traditional brick and mortar stores of any size can become highly digitized, and therefore cashier-less, which has massive potential for the future of the retail industry as a whole.
2、这些高度数字化的门店是如何运作的？
Jeff：基于人工智能技术，门店系统可以追踪顾客的逛店轨迹和拿放动作，从而创造部分自动化的零售环境。如此一来，顾客无需排队等候人工收银，也无需在自助收银台前频繁操作，挑选完商品直接走出闸机就能自动完成结算，整个购物流程更趋方便快捷。而这项技术也能革命性地降低运营成本，门店再也不用雇那么多店员，同时也能大幅降低盗损的风险以及由此造成的损失。
Jeff: At a high-level, this technology creates retail environments that are partially automated, with AI technology tracking customers movement and product selection. This empowers customers to quickly and conveniently purchase products without being checked out by a cashier, nor by using a self-checkout station. The technology also reduces costs for the retail business, who no longer have to employ so many people, while also drastically reducing the chance and cost, of theft.
3、传统门店也能快速转变为智慧商店吗？具体将变成什么样？
Jeff：得益于人工智能、物联网和大数据，线下传统门店也可以通过3D重建转变为智慧商店。这里使用到的人工智能技术，类似自动驾驶中常见的人工智能技术。
通过3D重建，店内的一切事物都将被数字化，包括货架、货架上的商品、店员以及顾客等等。因此，人工智能可以追踪到哪位顾客正在关注、拿放哪件商品，或者这位顾客是否喜欢某种特定品类，由此丰富用户画像，帮助经营方打造更个性化的消费体验、营销活动以及交互方式。而门店的全面数字化也意味着顾客不必再经历传统、繁琐的收银环节，人工智能可以自动追踪顾客所拿取的商品，完成结算并生成相应的订单。
Jeff: Through AI, IoT and big data, previously concrete stores can be converted into ‘smart stores’ via 3D reconstruction, an AI technology similar to that used in self-driving vehicles, but much more complex because LiDAR sensors - light detection and radar sensors - aren’t used.
Through 3D reconstruction, the physical store and the people within it are converted into a digital version where everything, including shelves, the products on those shelves, staff members and customers become a unique part of the digital landscape. This enables the AI to observe which customer is interacting with a certain product, or whether they favor a particular product category, providing specific insights on customer behavior which can be used to personalize their experience, marketing and the way the business communicates with them. Of course, this digitization also means that customers don’t have to go through the typical checkout and cashier experience, because the AI can track and bill them for the items they have selected without any interaction.
4、您认为有哪些因素可以加速此类转变？
Jeff：我认为当下的新冠疫情就是一个非常重要的催化因素，可能会加速零售业对人工智能技术的应用。由于需要保持社交距离以及非必要业务的暂停营业，各大零售企业都在寻找可以抵御此类全球性事件影响的方案。那么，智慧商店无疑就是这样一种方案，不仅可以避免顾客和店员在收银台前交互，也能追踪店内的顾客数量、彼此间的距离等，同时还将提醒店员对顾客高频光顾的区域进行消毒处理。
另一个催化因素则是人力成本。最近有很多新闻都提到了劳动力短缺的问题，想要从事服务行业的人越来越少，劳动力成本日益高昂。而我们的技术可以在减少人力需求的基础上保证门店的正常运营，这对零售业来说非常有吸引力。
Jeff: I believe we are currently experiencing a major catalyst which has the potential to accelerate the adoption of this technology: the COVID-19 pandemic. Thanks to social distancing mandates and forced closures of non-essential businesses, many retailers of all sizes are looking for ways to future-proof their businesses against similar types of global events. Smart stores can do this by not only eliminating interpersonal contact between cashiers and customers at PoS stations, but also by tracking the number of customers within a retail space, their vicinity to one another, and other actions within the store which can potentially alert an attendant to high-traffic areas that require additional sanitization.
Another catalyst could potentially be the cost of human labor. It’s growing day-by-day and fewer people want to occupy these jobs as evidenced by recent news about labor shortages. Technology like ours reduces the need for so many employees to keep a retail space functional, which is a major draw for any business.
5、您如何看待人工智能和机器学习技术在零售业的发展？最后是否会创造出100%自动化的商店？
Jeff：下一步应该不会激进到直接创建完全自动化的商店，尽管未来有一天这件事情可能会发生。相比之下，下一步这项技术更有可能着眼于落地场景的规模，从而可以适配任何大小的线下门店。零售业包含超市、小型超市、便利店等各种业态，但当前这项技术更适合规模较小的门店，所以我认为下一步是让这项技术能应用于不同的零售场景中去。事实上，我们的团队也一直在超市领域寻找创新的可能，希望尽快研发并推出相关的解决方案。
Jeff: I think the next step of this progression won’t be so drastic as creating an entirely automated store, although that could happen one day. Instead, the ultimate next step in the progression of this technology will be around scaling, to encompass stores of any size. If you think about retail currently, we have supermarkets, minimarts, convenience stores, micromarkets and more. This technology, as it stands, is more suited to smaller stores, so I think the next step is creating the same type of technology for different scenarios. In fact, our team has been looking at possible innovations within the supermarket sector, and we’re currently working towards these solutions - which I’m hoping we’ll be able to release relatively soon.
6、能否简要介绍一下云拿科技的发展历程？云拿科技是如何帮助传统门店快速转变为智慧商店的？
Jeff：云拿科技的一众创始人曾是校友，且在技术、零售、人工智能和机器学习领域有着相似的背景。2017年，大家注意到许多电商的线上市场陷入了增长瓶颈，希望通过开拓线下空间来实现产品服务的多元化。我们围绕这一需求开展了相关研发工作，并于2017年11月推出了首款设备，赋能线下零售，简化经营环节。从那时起，我们便逐渐与英特尔等一系列不同的客户和机构达成合作，在持续进行技术研发的同时将业务推广到多个国家和地区。现在，我们也已经发展成了智慧零售领域的领先者。
而在加速转型方面，我们不仅提供技术，也会与零售企业高管进行一系列交流，通过培训帮助他们改变传统的经营思维和流程，使之能高效地运营这些智慧商店。
Jeff: In 2017, our founders - we all went to school together and have similar backgrounds in technology, retail, AI and machine learning - noticed that many eCommerce companies were facing challenges with growth in the online space and were looking to add offline spaces to diversify their offerings. We took this idea and ran with it, developing and later launching our first wearable Cloudpick device in November 2017 to help empower and simplify offline retail. From there, we partnered with a range of different clients and organizations, including Intel, which helped us expand to a number of countries and regions while further developing our technology. Now, I’m fairly sure we have become the leading company in this specific space.
As for how we’re accelerating the transition; Not only are we providing the technology, but we also partner with C-level retail executives in a range of markets to ensure their transition in becoming a smart store is smooth and considered, helping them to launch through training to shift their processes so they can effectively operate these new types of stores.
关于 Retail IT Insights
Retail IT Insights致力于为零售业提供重要、创新的技术解决方案资源，帮助从业者在各个经营环节高效决策。除此之外还提供如何让零售商通过使用前沿解决方案来实现关键业务目标的深刻洞察，包括媒体新闻、网络研讨会以及在其他地方无法获取的独家内容等等。目前，网站每年访问人数高达114万，每年浏览量高达563万。