当Curtis Carver博士在2015年接任阿拉巴马大学伯明翰分校的首席信息官时，他遇到了一个他称之为“计算机科学博物馆”的地方——过去30年里市场上所有操作系统、存储设备和应用程序的实例。
Carver当时的首要任务是清理过时的混乱局面，将整个技术基础设施合理化，整合到一个单一、全面的平台上，使大学的数千名科学家能够进行创新，并取得突破性的研究发现。“我们开始了一种合理化的模式，那就是混合云。当你的研究人员在进行个性化医疗或基因组学研究时，数据集是如此之大，以至于你的网络有多快都无关紧要。光速成为了一个限制因素。”
七年后，Carver现在负责管理混合云基础设施，主要基于微软Azure和Kubernetes, Tableau和微软Power BI用于分析，微软的Azure人工智能框架，以及大量Nvidia图形处理器，所有这些都通过100千兆连接连接到附近私有数据中心的高性能计算机集群。
看来他的现代化方式正在产生影响。去年秋天，Carver所说的该州医学研究中心UAB在基因组学方面取得了一项发现，可以将猪肾脏移植到人类身上，这一潜在的拯救生命的突破引起了CNN和全球主要媒体的关注。
UAB医疗团队将两个猪肾移植到一个脑死亡的人体内，目前正在计划将猪肾移植到活人体内的临床试验，这一努力将需要政府的批准。如果成功，该手术将对肾衰竭死亡率产生深远影响，肾衰竭每年夺去数十万人的生命。
Carver并不认为这是他的功劳。但UAB正在进行的数字化转型和随后的科学胜利，无疑是推动其基因组学研究和筹集资金以实现更多突破的一个因素。举个例子:该大学刚刚破土动工，斥资7500万美元建造两栋大楼，作为UAB基因组学研究中心。
一、大型制药公司正以前所未有的方式投资于技术，它们的首席信息官们应该采用前沿技术来保持领先地位，这是至关重要的。
除了彻底改造整个IT基础设施，Carver还引导UAB的IT员工投入更多资源，以实现科学创新。如今，在他的IT员工中，只有不到一半的人被派去“维持运转”，比如解决帮助台问题，而更大比例的人被派去解决基因组学和医学科学家的下一代技术需求。
这种肌肉对科学家来说很重要，他们曾经在所有IT领域——存储、电子邮件和处理能力都遇到了限制。Carver的目标是为UAB的研究人员提供无限的IT服务，并且正在朝着这个方向努力。
这位首席信息官表示:“我们最近的采购集中在专门为神经网络和人工智能构建的高性能计算处理器上，我们在基因组医学中大量使用这些处理器。对于我们的基因组学来说，这些处理器是量身定制的。”
UAB每年约8.5亿美元的赞助研究也有助于提升IT所能提供的服务。
Carver说:“整个想法是，我们从一个稀缺的时代走向一个富足的时代。你可以管理IT和预算，让人们能够专注于胰腺癌。我们的想法是创造一套人们可以实际使用的工具，而不是一钱不值的东西。”
IDC公司生命科学研发战略和技术研究副总裁、分析师Nimita Limaye博士表示，考虑到来自私营部门的竞争，像UAB这样的机构在IT方面的努力至关重要。
她说:“大型制药公司正以前所未有的方式投资于技术，并积极与人工智能供应商建立合作关系。大学和学术医疗中心需要加强它们作为生命科学行业创新生态系统的地位，因此，它们的首席信息官们应该采用前沿技术来保持领先地位，这是至关重要的。”
二、我们希望建立一个数字市场并进行部署，而不是通过承包工具来解决一个常见问题。
Carver表示，为了实现这一目标，UAB的下一步是围绕扩大其人工智能和数据分析平台做出重大决策。Carver并非独自负责长期规划。UAB有两个委员会，数据治理和数字战略，他们的工作将有助于指导决策，例如，采用哪个下一代人工智能平台和数据湖。
作为其数据治理计划的一部分，UAB寻求在核心数据系统中建立“深厚的专业知识”，包括数据字典、数字验证器和数字市场。Carver表示，UAB“倾向于”使用ServiceNow，但也在考虑将Salesforce作为其数字市场。
“(我们希望)建立一个数字市场并进行部署，而不是通过承包工具来解决一个常见问题。你可以解决整个大学的(采购)问题，而不是一个小单位，”Carver说。“这是在促进创新，但我们这样做的方式是，我们正在创造竞争优势和共同的方法，并利用我们作为一所价值数十亿美元的大学所能带来的全面规模经济。”
Carver说，UAB是微软的大客户，但也与亚马逊和谷歌签订了主服务协议。例如，该大学的学习管理系统运行在亚马逊网络服务上，其他服务托管在谷歌云平台上。
UAB目前在本地运行一个Tableau服务器，许多行政部门和一些研究人员都在使用。该校的一位女发言人说，该校的一些院系也在使用Microsoft Power BI。
UAB计划在未来实现一个集中的Power BI服务器，并开发一个集中的报告目录，该目录将显示在Tableau、Power BI和其他分析解决方案中生成的报告。她补充说，这是为了帮助用户更快地找到报告和相关数据。
在人工智能领域，UAB一直在使用微软的自然语言处理技术和Azure AI框架。例如，Carver的团队为不上课的学生开发了“侵入式建议代理”，以及回答学生和教职员工常见问题的聊天机器人。
但在人工智能、自动化、数据分析以及卡弗钟爱的项目之一——安全方面，还有很多工作要做。他说，UAB的数字化转型仍然是一项多年的努力，不仅将主导2022年和2023年，而且将主导可预见的未来。但支持该大学研究工作的工作是值得的。
“人类所能做的是了不起的。电脑能做的事太棒了。但当你把两者结合起来，当你让人类和计算机相互作用，共同解决问题时，这就更好了，”Carver说。“这不是一个，也不是另一个。但我们如何提供协同工作的系统呢?”
Carver旨在为科学家提供一个非常灵活的平台，使他们能够在15分钟内访问自定义级别的处理能力、存储和网络。“让他们专注于科学。我们会处理事情的后端，”Carver说。“因此，他们将赢得更多的资助，进行科学发现，并改变世界。”
原文：
When Dr. Curtis Carver, Ph.D., took over as CIO at University of Alabama-Birmingham in 2015, he confronted a “computer science museum,” as he calls it — instances of every operating system, storage device, and application on the market for the past 30 years.
Carver’s first priority then was to clean up the outdated mess and rationalize the entire technology infrastructure onto a single, comprehensive platform that would empower the university’s thousands of scientists to innovate and make breakthrough research discoveries.
“We started a movement towards a rationalized model, and that was a hybrid cloud,” Carver tells CIO.com. “When your researchers are doing personalized medicine, or genomics research, the datasets are so large that it doesn’t matter how fast your network is. It is the speed of light that becomes a limiting factor.”
Seven years later, Carver now oversees a hybrid cloud infrastructure based primarily on Microsoft Azure and Kubernetes, Tableau and Microsoft Power BI for analytics, Microsoft’s Azure AI framework, and lots of Nvidia graphics processors, all connected to a high-performance computer cluster at a nearby private data center via a 100 Gigabit connection.
It appears his modernization approach is having an impact.
Last fall, the UAB, which Carver says is the medical research hub of the state, made a genomics discovery that enables pig kidneys to be transplanted into humans — a potential life-saving breakthrough that caught the attention of CNN and major media outlets globally.
The UAB medical team transplanted two pig kidneys into a brain-dead human recipient and is currently planning a clinical trial to transplant the pig kidneys into living humans — an endeavor that will require approvals from the government. If successful, the procedure could have a profound impact on kidney failure mortality, which claims hundreds of thousands of lives each year.
Carver claims no credit for any of that. But UAB’s ongoing digital transformation and subsequent scientific win is no doubt a factor in enabling its genomics research and generating the kind of fundraising that leads to more breakthroughs. Case in point: The university just broke ground on a $75 million project to construct two buildings that will serve as the UAB Genomics Research Hub.
Transforming to support scientific innovation
In addition to overhauling the entire IT infrastructure, Carver pivoted UAB’s IT workforce to put more resources into enabling scientific innovation. Today, less than half of his IT staff is deployed to “keep the lights on” matters such as addressing helpdesk issues while a larger percentage is tackling the next-generation technology needs of its genomics and medical scientists.
That kind of muscle is important to scientists, who once were bumping up against limits in all IT areas — storage, e-mail, and processing power. Carver aims to offer unlimited IT services to UAB’s researchers and is well on his way.
“Our most recent purchases have been focused on high-performance computing processors built specifically for neural networks and artificial intelligence, which we use quite a bit in genomics medicine,” the CIO says. “For our genomics, those processors are well tailored for that.”
The roughly $850 million in sponsored research UAB brings in each year is also helping to elevate what IT can provide.
“The whole idea is we move from an era of scarcity to an era of abundance,” Carver says. “And you can manage the IT and budget in such a way that people are empowered to focus on pancreatic cancer. … The idea is to create a palette of tools that people can actually use rather than nickel and diming it.”
And analyst Dr. Nimita Limaye, research vice president of life sciences R&D strategy and technology at IDC, says such IT efforts at institutions like UAB are vital given competition from the private sector.
“Big pharma is investing in technology like never before and it is aggressively forging partnerships with AI vendors,” she says. “Universities and academic medical centers need to reinforce their positions as innovation ecosystems for the life sciences industry, and therefore, it is critical that their CIOs should adopt cutting-edge technologies to stay ahead of the curve.”
Next up: AI and data lake decisions
To that end, UAB’s next step is to tackle big decisions around expanding its AI and data analytics platforms, says Carver, who is not handling the long-term planning alone. UAB has two committees, Data Governance and Digital Strategy, whose work will help guide decisions about which next-generation AI platform and data lake to adopt, for instance.
As part of its data governance plan, UAB seeks to build “deep expertise” in core data systems, including a data dictionary, digital validators, and a digital marketplace. Carver says UAB is “leaning toward” using ServiceNow but is also considering Salesforce for its digital marketplace.
“Rather than going through a contracting vehicle to solve a common problem, [we want to build] a digital marketplace and deploy it. You can solve [purchasing] for the entire university instead of one small unit,” Carver says. “It’s fostering innovation, but we’re doing it in such a way that we’re creating competitive advantage and a common approach and leveraging the full economies of scale that we can bring as a multibillion-dollar university.”
UAB is a big Microsoft customer but also has master service agreements with Amazon and Google, Carver says. The university’s learning management system, for example, runs on Amazon Web Services, and other services are hosted on Google Cloud Platform.
UAB currently runs a Tableau server on-premises that many administrative departments and some researchers use. A few departments on campus also use Microsoft Power BI, a UAB spokeswoman says.
UAB plans to implement a centralized Power BI server in the future and develop a centralized reporting catalog that will show reports produced in Tableau, Power BI, and other analytics solutions. This is aimed at helping users find reports and the associated data more quickly, she adds.
On the artificial intelligence front, UAB has been using Microsoft’s natural language processing technology and Azure AI frameworks. For example, Carver’s team has developed “intrusive advising agents” for students who are not attending classes, and chatbots that answer common questions from students and staff.
But so much more has to be done in AI, automation, and data analytics, as well as security, one of Carver’s pet projects. UAB’s digital transformation remains a multi-year endeavor that’s going to dominate not only 2022 and 2023 but the foreseeable future, he says. But the work in support of the university’s research endeavors is rewarding.
“What humans can do is awesome. What computers can do is awesome. But when you combine the two, when you have a human and computer interacting together to jointly solve a problem, it’s way better,” Carver says. “It’s not one, or the other. But how do we provide systems that work in synergy.”
Carver aims to provide scientists with a very agile platform that gives them access to custom levels of processing power, storage, and networking in 15 minutes. “Let them focus on the science. We’ll handle the back end of things,” Carver says. “So, they’ll win more grants, make scientific discoveries, and change the world.”
本文由数字化转型网（www.szhzxw.cn）翻译而成，作者：Paula Rooney；翻译：数字化转型网郑亚茹；翻译审核：数字化转型网默然。
免责声明: 本网站(https://www.szhzxw.cn/)内容主要来自原创、合作媒体供稿和第三方投稿，凡在本网站出现的信息，均仅供参考。本网站将尽力确保所提供信息的准确性及可靠性，但不保证有关资料的准确性及可靠性，读者在使用前请进一步核实，并对任何自主决定的行为负责。本网站对有关资料所引致的错误、不确或遗漏，概不负任何法律责任。
本网站刊载的所有内容(包括但不仅限文字、图片、LOGO、音频、视频、软件、程序等) 版权归原作者所有。任何单位或个人认为本网站中的内容可能涉嫌侵犯其知识产权或存在不实内容时，请及时通知本站，予以删除。
